Qwen2.5-VL（开源视觉模型） 2.5

HF_ENDPOINT=https://hf-mirror.com huggingface-cli download Qwen/Qwen2.5-VL-7B-Instruct --local-dir ./qwen2.5-vl-7b

云端 API：阿里云百炼提供 qwen-vl-plus / qwen-vl-max。

from transformers import Qwen2VLForConditionalGeneration, AutoTokenizer
from transformers import Qwen2VLProcessor
import torch

model = Qwen2VLForConditionalGeneration.from_pretrained("./qwen2.5-vl-7b", device_map="auto")
processor = Qwen2VLProcessor.from_pretrained("./qwen2.5-vl-7b")

from PIL import Image
img = Image.open("cat.jpg")
prompt = "<image>\n这张图片里有什么？"
inputs = processor(text=prompt, images=img, return_tensors="pt").to("cuda")
out = model.generate(**inputs, max_new_tokens=128)
print(processor.decode(out[0], skip_special_tokens=True))

Ollama 一条命令：

ollama pull qwen2.5vl
ollama run qwen2.5vl

Transformers 推理：

from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor
from PIL import Image

model = Qwen2_5_VLForConditionalGeneration.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct", device_map="auto")
processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct")
img = Image.open("demo.png").convert("RGB")
msgs = [{"role": "user", "content": [
    {"type": "image"},
    {"type": "text", "text": "这张图里有什么？"},
]}]
text = processor.apply_chat_template(msgs, tokenize=False, add_generation_prompt=True)
inputs = processor(text=[text], images=[img], return_tensors="pt")
out = model.generate(**processor.to_hf_device_map(inputs), max_new_tokens=128)
print(processor.batch_decode(out[:, inputs.input_ids.shape[1]:], skip_special_tokens=True)[0])