TrishanuDas/tayavision-multilingual

TayaVision - Tiny Aya Vision (Instruct)

This model is an instruct sweep on liuhaotian/LLaVA-Instruct-150K with lora_a_lr_multiplier=1 and lora_b_lr_multiplier=0.1.

import torch
from PIL import Image
from transformers import AutoModelForCausalLM, AutoProcessor

repo = "TrishanuDas/tayavision-multilingual"

model = AutoModelForCausalLM.from_pretrained(repo, torch_dtype=torch.bfloat16, trust_remote_code=True)
model = model.to("cuda").eval()

processor = AutoProcessor.from_pretrained(repo, trust_remote_code=True)

image = Image.open("your_image.jpg").convert("RGB")

messages = [
    {"role": "user", "content": [
        {"type": "image"},
        {"type": "text", "text": "Describe this image in detail."},
    ]},
]

inputs = processor.apply_chat_template(
    messages,
    images=image,
    add_generation_prompt=True,
    tokenize=True,
    return_tensors="pt",
)
inputs = {key: value.to("cuda") for key, value in inputs.items()}

with torch.inference_mode():
    output_ids = model.generate(
        **inputs,
        max_new_tokens=256,
        do_sample=False,
        use_cache=True,
    )

response = processor.tokenizer.decode(
    output_ids[0, inputs["input_ids"].shape[1]:],
    skip_special_tokens=True,
)
print(response)