# Load model directly
from transformers import AutoProcessor, AutoModelForVisualQuestionAnswering
processor = AutoProcessor.from_pretrained("kkatiz/THAI-BLIP-2")
model = AutoModelForVisualQuestionAnswering.from_pretrained("kkatiz/THAI-BLIP-2")Quick Links
THAI-BLIP-2
fine-tuned for image captioning task from blip2-opt-2.7b-coco with MSCOCO2017 thai caption.
How to use:
from transformers import Blip2ForConditionalGeneration, Blip2Processor
from PIL import Image
import torch
device = "cuda" if torch.cuda.is_available() else "cpu"
processor = Blip2Processor.from_pretrained("kkatiz/THAI-BLIP-2")
model = Blip2ForConditionalGeneration.from_pretrained("kkatiz/THAI-BLIP-2", device_map=device, torch_dtype=torch.bfloat16)
img = Image.open("Your image...")
inputs = processor(images=img, return_tensors="pt").to(device, torch.bfloat16)
# Adjust your `max_length`
generated_ids = model.generate(**inputs, max_length=20)
generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)
print(generated_text)
- Downloads last month
- 84
Model tree for kkatiz/THAI-BLIP-2
Base model
Salesforce/blip2-opt-2.7b-coco
# Use a pipeline as a high-level helper # Warning: Pipeline type "image-to-text" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("image-to-text", model="kkatiz/THAI-BLIP-2")