Views
No views yet
1from optimum.intel import OVModelForVisualCausalLM
2from transformers import AutoProcessor
3from PIL import Image
4
5model_id = "CelesteImperia/MiniCPM-V-2.6-OpenVINO-INT8"
6processor = AutoProcessor.from_pretrained(model_id, trust_remote_code=True)
7model = OVModelForVisualCausalLM.from_pretrained(model_id, trust_remote_code=True)
8
9image = Image.open("path/to/your/image.jpg")
10prompt = "Perform a detailed audit of this visual scene."
11
12inputs = processor(text=[prompt], images=[image], return_tensors="pt")
13outputs = model.generate(**inputs, max_new_tokens=256)
14print(processor.decode(outputs[0], skip_special_tokens=True))VLMPipeline in the OpenVINO.GenAI framework, enabling high-precision visual automation in industrial C# environments.1using OpenVino.GenAI;
2
3// 1. Initialize the Visual-LLM Pipeline
4var device = "CPU"; // Leverage "GPU" for accelerated dual-GPU inference
5using var pipe = new VLMPipeline("path/to/minicpm-v-int8-model", device);
6
7// 2. Load Visual Input
8var image = OpenVino.GenAI.Utils.LoadImage("industrial_capture.jpg");
9var prompt = "What are the specific technical identifiers visible on this component?";
10
11// 3. Multimodal Execution
12var result = pipe.Generate(prompt, image);
13Console.WriteLine(result.Texts[0]);| Platform | Support Link |
|---|---|
| Global & India | Support via Razorpay |
