---
library_name: transformers
license: apache-2.0
datasets:
- merve/vqav2-small
---
**Note**: There's a [newer version of this repository](https://huggingface.co/merve/idefics3-llama-vqav2), and I have uploaded training script to smol-vision along with new notebook! 

![image/png](https://cdn-uploads.huggingface.co/production/uploads/6141a88b3a0ec78603c9e784/PebmPLcCig5BlpUS99VUc.png)

# Idefics3Llama Fine-tuned using QLoRA on VQAv2

- This is the [Idefics3Llama](https://huggingface.co/HuggingFaceM4/Idefics3-8B-Llama3) model QLoRA fine-tuned on a very small part of [VQAv2](https://huggingface.co/datasets/merve/vqav2-small) dataset.

- Find the fine-tuning notebook [here](https://github.com/merveenoyan/smol-vision/blob/main/Idefics_FT.ipynb). 

## Usage

You can load and use this model as follows.

```python

from transformers import Idefics3ForConditionalGeneration, AutoProcessor

peft_model_id = "merve/idefics3llama-vqav2"
base_model_id = "HuggingFaceM4/Idefics3-8B-Llama3"
processor = AutoProcessor.from_pretrained(base_model_id)
model = Idefics3ForConditionalGeneration.from_pretrained(base_model_id)
model.load_adapter(peft_model_id).to("cuda")

```

This model was conditioned on a prompt "Answer briefly.".

```python
from PIL import Image
import requests
from transformers.image_utils import load_image

DEVICE = "cuda"

image = load_image("https://huggingface.co/spaces/merve/OWLSAM2/resolve/main/buddha.JPG")


messages = [
          {
              "role": "user",
              "content": [
                  {"type": "text", "text": "Answer briefly."},
                  {"type": "image"},
                  {"type": "text", "text": "Which country is this located in?"}
              ]
          }
      ]

text = processor.apply_chat_template(messages, add_generation_prompt=True)
inputs = processor(text=text, images=image, return_tensors="pt", padding=True).to("cuda")
```

We can infer.

```python
generated_ids = model.generate(**inputs, max_new_tokens=500)
generated_texts = processor.batch_decode(generated_ids, skip_special_tokens=True)
print(generated_texts)

##['User: Answer briefly.<row_1_col_1><row_1_col_2><row_1_col_3><row_1_col_4>\n<row_2_col_1>
# <row_2_col_2><row_2_col_3><row_2_col_4>\n<row_3_col_1><row_3_col_2><row_3_col_3>
# <row_3_col_4>\n\n<global-img>Which country is this located in?\nAssistant: thailand\nAssistant: thailand']
```