https://github.com/four43/ml-fast-ai
Machine Learning training via fast.ai
https://github.com/four43/ml-fast-ai
Last synced: 8 months ago
JSON representation
Machine Learning training via fast.ai
- Host: GitHub
- URL: https://github.com/four43/ml-fast-ai
- Owner: four43
- Created: 2018-01-05T20:41:10.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2018-01-08T05:21:22.000Z (over 8 years ago)
- Last Synced: 2025-08-08T04:33:40.643Z (10 months ago)
- Language: Jupyter Notebook
- Size: 22.1 MB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Machine Learning with [fast.ai](https://fast.ai)
## Setup
I am running most of this locally on a workstation with an nVidia GPU (currently a GTX1080). We can use nvidia-docker to keep a clean environment but we will still need to ensure we have the proper setup on the host machine. These [instructions](https://github.com/fastai/courses/blob/master/setup/install-gpu.sh) are great to setup the environment and are required to get running before nvidia-docker. See `host-setup.sh`
## Running
Ensure `nvidia-docker` is setup and working. See `host-setup.sh` to install it.
GPU Test/Information:
```bash
docker run \
--runtime=nvidia \
--rm \
nvidia/cuda nvidia-smi
```
Run our notebook service:
```bash
docker build -t ml-fast-ai .
docker run \
--runtime=nvidia \
--rm \
--name ml-fast-ai \
--network="host" \
-v $(pwd)/course-content:/data \
-p 8888:8888 \
ml-fast-ai
```
Monitor GPU Usage:
```bash
docker run \
-ti \
--runtime=nvidia \
--rm \
nvidia/cuda:8.0-cudnn7-runtime-ubuntu16.04 \
/bin/bash -c "watch -n 1 nvidia-smi"
```
## Practical Deep Learning for Coders - Part 1