This is the repository for the ECS189G course on Deep Learning. The main purpose of this project is to help students **get familiar** with classic deep learning models and **use them to solve** concrete real-world problems.
I'm a bit cautious about changing the flow from how we've been running things in the terminal. I would have a preference for a script being uploaded, just because we have everything version controlled + linted, etc. Plus, a separate task will involve running jobs in this remote execution environment. Are there solutions that check in our code from a repo and run a job, or maybe we invoke a command from the cli
Is it possible for someone to look into GCP/AWS/Azure to see if we can get credits for training? This is a good place to look into costs. Realistically we can do a trial and use credits
I'm a bit cautious about changing the flow from how we've been running things in the terminal. I would have a preference for a script being uploaded, just because we have everything version controlled + linted, etc. Plus, a separate task will involve running jobs in this remote execution environment. Are there solutions that check in our code from a repo and run a job, or maybe we invoke a command from the cli
Is it possible for someone to look into GCP/AWS/Azure to see if we can get credits for training? This is a good place to look into costs. Realistically we can do a trial and use credits
Colab
Saturn
kaggle notebooks