This project demonstrates fine-tuning Vision-Language Models (VLMs) for image understanding tasks, specifically using the Qwen2.5-VL-7B model for wildfire detection from satellite imagery using GRPO (Generalized Reward Preference Optimization).
## Overview
The project includes:
- **Interactive Training Interface**: Streamlit-based UI for configuring and monitoring VLM fine-tuning
- **GRPO Training**: Advanced preference optimization for better reasoning capabilities
> **Note**: These instructions assume you are already inside the Docker container. For container setup, refer to the main project README at `vlm-finetuning/assets/README.md`.
If you already have a fine-tuned checkpoint, place it in the `saved_model/` folder. Your directory structure should look something like this. Note that your checkpoint number can be different.
If you already have a finetuned checkpoint that you would like to just use for a comparative analysis against the base model, skip directly to the [Finetuned Model Inference](#5-finetuned-model-inference) section.
The project uses a **Wildfire Detection Dataset** with satellite imagery for training the model to identify wildfire-affected regions. The dataset includes:
For this finetuning playbook, we will use the [Wildfire Prediction Dataset](https://www.kaggle.com/datasets/abdelghaniaaba/wildfire-prediction-dataset) from Kaggle. Visit the kaggle dataset page [here](https://www.kaggle.com/datasets/abdelghaniaaba/wildfire-prediction-dataset) to click the download button. Select the `cURL` option in the `Download Via` dropdown and copy the curl command.
Before we start finetuning, let's start spin up the demo UI to evaluate the base model's performance on this task.
### 3.1 Spin up the Streamlit demo
```bash
streamlit run Image_VLM.py
```
Access the streamlit demo at http://localhost:8501/.
### 3.2 Wait for demo spin-up
When you access the streamlit demo for the first time, the backend triggers vLLM servers to spin up for the base model. You will see a spinner on the demo site as vLLM is being brought up for optimized inference. This step can take upto 15 mins.
After the streamlit demo is fully loaded, you should be able to see a similar UI state that is ready for inference.
Since we are currently focused on inferring the base model, let's scroll down to the `Image Inference` section of the UI. Here, you should see a sample pre-loaded satellite image of a potentially wildfire-affected region.
Enter your prompt in the chat box and hit `Generate`. Your prompt would be first sent to the base model and you should see the generation response on the left chat box. If you did not provide a finetuned model, you should not see any generations from the right chat box.
As you can see, the base model is incapable of providing the right response for this domain-specific task. Let's try to improve the model's accuracy by performing GRPO finetuning.
We will perform GRPO finetuning to add reasoning capabilities to our base model and improve the model's understanding to the underlying domain. Considering that you have already spun up the streamlit demo, scroll to the `GRPO Training section`.
You can additionally choose whether the layers you want to finetune in the VLM. For the best performance, ensure that all options are toggled on. Note that this will increase the model training time as well.
After configuring all the parameters, hit `Start Finetuning` to begin the training process. You will need to wait about 15 mins for the model to load and start recording metadata on the UI. As the training progresses, information such as the loss, step, and GRPO rewards will be recorded on a live table.
The default loaded configuration should give you reasonable accuracy, taking 100 steps of training over a period of upto 2 hours. We achieved our best accuracies with around 1000 steps of training, taking close to 16 hours.
After training is complete, the script automatically merges lora weights into the base model. After the training process has reached the desired number of training steps, it can take 5 mins to merge the lora weights.
If you wish to stop training, just hit the `Stop Finetuning` button. Please use this button only to interrupt training. This button does not guarantee that the checkpoints will be properly stored or merged with lora adapter layers.
Once you stop training, the UI will automatically bring up the vLLM servers for the base model and the newly finetuned model.
## 5. Finetuned Model Inference
Now we are ready to perform a comparative analysis between the base model and the finetuned model.
### 5.1 (Optional) Spin up the Streamlit demo
If you haven't spun up the streamlit demo already, execute the following command. If had just just stopped training and are still within the live UI, skip to the next step.
```bash
streamlit run Image_VLM.py
```
Access the streamlit demo at http://localhost:8501/.
### 5.2 vLLM startup
Regardless of whether you just spun up the demo or just stopped training, please wait about 15 mins for the vLLM servers to be brought up.
### 5.3 Run finetuned model inference
Scroll down to the `Image Inference` section, and enter your prompt in the provided chat box. Upon clicking `Generate`, your prompt would be first sent to the base model and then to the finetuned model. You can use the following prompt to quickly test inference
`Identify if this region has been affected by a wildfire`
If you trained your model sufficiently enough, you should see that the finetuned model is able to perform reasoning and provide a concise, accurate answer to the prompt. The reasoning steps are provided in the markdown format, while the final answer is bolded and provided at the end of the model's response.
If you are facing VRAM issues where the model fails to load or offloads to cpu/meta device, ensure you bring down all docker containers and flush out dangling memory.
```bash
docker ps
docker rm <CONTAINER_ID_1>
docker rm <CONTAINER_ID_2>
docker system prune
sudo sh -c 'sync; echo 3 > /proc/sys/vm/drop_caches'
