24
Jan
2020
/
Carey Phelps, Product Lead

Visualizing TensorFlow 2 models with Weights & Biases

It's easy to integrate your TensorFlow models with Weights & Biases. With this quick integration you can see your live metrics streaming in to our visualizations, and compare new results to your previous baselines.

Try the Colab → to train a CNN, or this one → to train a Perceptron.

In one minute you can train these TensorFlow 2 models and see live results → streaming into a Weights & Biases project.

Integrating with Weights & Biases

To visualize your own models, add a few lines of code to any TensorFlow script to start seeing results:

  1. import wandb at the top of your script
  2. wandb.init(config=param_dict) initialize a new run and pass in a dictionary of the model's hyperparameters
  3. wandb.log({"loss": loss, "val_acc": val_acc}) log output metrics to see them graphed over time


Next Steps

To see your live results in Weights & Biases:

  1. Get a free personal account. Sign up →
  2. Run the hosted notebook. Google Colab →
  3. See results appear in this W&B report →

Newsletter

Enter your email to get updates about new features and blog posts.

Weights & Biases

We're building lightweight, flexible experiment tracking tools for deep learning. Add a couple of lines to your python script, and we'll keep track of your hyperparameters and output metrics, making it easy to compare runs and see the whole history of your progress. Think of us like GitHub for deep learning.

Partner Program

We are building our library of deep learning articles, and we're delighted to feature the work of community members. Contact Lavanya to learn about opportunities to share your research and insights.

Try our free tools for experiment tracking →