Apple’s FastVLM is here and it is currently emerging as a game-changer in Vision Language Models (VLMs), combining fast speed with state-of-the-art accuracy. It is designed for real-time performance without compromising on output quality, and uses FastViTHD, a hybrid vision encoder that drastically cuts down the number of image tokens, making inference lightning-fast even with high-resolution inputs. Its smallest variant outpaces LLaVA-OneVision-0.5B with an outstanding 85x faster Time-to-First-Token (TTFT), all while being 3.4x smaller in size. On the other end of the scale, its larger models powered by Qwen2-7B surpass recent competitors like Cambrian-1-8B, achieving 7.9x faster TTFT using just a single image encoder. If you’re running models on powerful servers or experimenting on mobile devices, FastVLM delivers top-tier vision-language capabilities with low latency and unmatched efficiency.
In this guide, we’re going to cover step-by-step process to install and run FastVLM locally or on cloud.
Prerequisites
The minimum system requirements for running this model are:
Step-by-step process to install and run FastVLM by Apple
For the purpose of this tutorial, we’ll use a GPU-powered Virtual Machine by NodeShift since it provides high compute Virtual Machines at a very affordable cost on a scale that meets GDPR, SOC2, and ISO27001 requirements. Also, it offers an intuitive and user-friendly interface, making it easier for beginners to get started with Cloud deployments. However, feel free to use any cloud provider of your choice and follow the same steps for the rest of the tutorial.
Step 1: Setting up a NodeShift Account
Visit app.nodeshift.com and create an account by filling in basic details, or continue signing up with your Google/GitHub account.
If you already have an account, login straight to your dashboard.
Step 2: Create a GPU Node
After accessing your account, you should see a dashboard (see image), now:
- Navigate to the menu on the left side.
- Click on the GPU Nodes option.
- Click on Start to start creating your very first GPU node.
These GPU nodes are GPU-powered virtual machines by NodeShift. These nodes are highly customizable and let you control different environmental configurations for GPUs ranging from H100s to A100s, CPUs, RAM, and storage, according to your needs.
Step 3: Selecting configuration for GPU (model, region, storage)
- For this tutorial, we’ll be using 1x RTX A6000 GPU, however, you can choose any GPU as per the prerequisites.
- Similarly, we’ll opt for 200GB storage by sliding the bar. You can also select the region where you want your GPU to reside from the available ones.
Step 4: Choose GPU Configuration and Authentication method
- After selecting your required configuration options, you’ll see the available GPU nodes in your region and according to (or very close to) your configuration. In our case, we’ll choose a 1x RTX A6000 48GB GPU node with 64vCPUs/63GB RAM/200GB SSD.
2. Next, you’ll need to select an authentication method. Two methods are available: Password and SSH Key. We recommend using SSH keys, as they are a more secure option. To create one, head over to our official documentation.
Step 5: Choose an Image
The final step is to choose an image for the VM, which in our case is Nvidia Cuda.
That’s it! You are now ready to deploy the node. Finalize the configuration summary, and if it looks good, click Create to deploy the node.
Step 6: Connect to active Compute Node using SSH
- As soon as you create the node, it will be deployed in a few seconds or a minute. Once deployed, you will see a status Running in green, meaning that our Compute node is ready to use!
- Once your GPU shows this status, navigate to the three dots on the right, click on Connect with SSH, and copy the SSH details that appear.
As you copy the details, follow the below steps to connect to the running GPU VM via SSH:
- Open your terminal, paste the SSH command, and run it.
2. In some cases, your terminal may take your consent before connecting. Enter ‘yes’.
3. A prompt will request a password. Type the SSH password, and you should be connected.
Output:
Next, If you want to check the GPU details, run the following command in the terminal:
!nvidia-smi
Step 7: Set up the project environment with dependencies
- Create a virtual environment using Anaconda.
conda create -n fastvlm python=3.10 -y && conda activate fastvlm
Output:
2. Clone the official repository of Apple-FastVLM.
git clone https://github.com/apple/ml-fastvlm.git
Output:
4. Move inside the project directory and run the following command to install the current directory as a package in editable mode.
cd ml-fastvlm
pip install -e .
Output:
Step 8: Download and Run the model
- Download all model checkpoints.
chmod a+x get_models.sh
bash get_models.sh
Output:
2. Once all the checkpoints are downloaded, we’ll connect our local VSCode editor to the remote server to do some changes in the python script before inference.
If you’re using a GPU through a remote server (e.g., NodeShift), you can connect it to your visual studio code editor by following the steps below:
a) Install the “Remote-SSH” Extension by Microsoft on VS Code.
b) Type “Remote-SSH: Connect to Host” on the Command Palette.
c) Click on “Add a new host”.
d) Enter the host details, such as username and SSH password, and you should be connected.
3. Open predict.py and replace all the instances of "mps" with "cuda" if running with GPU and with "cpu" if running with CPU only. If you’re running this model on your local Mac machine, then you do not need to replace anything.
4. Once done, upload images that you want to test your model on. Below are some results that this model generated for two different tasks.
Use the given commands to run inference:
For describing image:
(Replace /path/to/checkpoint-dir with the full path to the exact model that you want to use for inference. E.g. llava-fastvithd_7b_stage3. Also replace /path/to/image.png with the full path to the place where you have uploaded your images.)
python predict.py --model-path /path/to/checkpoint-dir \
--image-file /path/to/image.png \
--prompt "Describe the image."
Here’s a blurry image that we’re using in our prompt:
Output:
For extracting text from image:
python predict.py --model-path /path/to/checkpoint-dir \
--image-file /path/to/image.png \
--prompt "Extract the text from this image."
Here’s the output we received for the given image:
Image:
Output:
Conclusion
Apple’s FastVLM stands out as a transformative Vision Language Model, delivering exceptional speed and accuracy through its innovative FastViTHD encoder. By significantly reducing visual token counts and encoding latency, FastVLM ensures real-time performance even with high-resolution images, making it ideal for diverse applications ranging from mobile devices to large-scale cloud deployments. To fully leverage FastVLM’s capabilities, deploying it on a robust and scalable infrastructure is crucial. NodeShift cloud offers an optimal solution by providing GPU-powered Virtual Machines that are both affordable and compliant with industry standards like GDPR, SOC2, and ISO27001. With NodeShift’s user-friendly platform, developers can effortlessly set up and run FastVLM, ensuring seamless integration and efficient performance across various environments.
