https://github.com/mindee/tf-builds
https://github.com/mindee/tf-builds
Last synced: 9 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/mindee/tf-builds
- Owner: mindee
- Created: 2019-05-23T16:43:43.000Z (about 7 years ago)
- Default Branch: master
- Last Pushed: 2021-03-21T16:30:04.000Z (over 5 years ago)
- Last Synced: 2025-10-06T03:08:46.959Z (9 months ago)
- Size: 1.95 KB
- Stars: 5
- Watchers: 0
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# tf-builds
This is a list of optimized WHL Tensorflow builds for cloud computing use with Intel® Xeon® Scalable CPUs.
## Download links
|Date of build|Tensorflow|Python|Platform|Architecture|Links|
|-------------|----------|------|--------|------------|-----|
|2019-07-29|1.14.0|3.6|Linux|skylake|[download](https://github.com/publicMindee/tf-builds/releases/download/1.14.0-no-mkl-c5/tensorflow-1.14.0-cp36-cp36m-linux_x86_64.whl)|
## Installation
In your Python environment, run
```pip install --upgrade /path/to/file.whl```
## Best VMs
A non exhaustive list of VMs that work best for CPU Tensorflow inference.
#### AWS
> *Intel® Xeon® Platinum 8124M*
* c5.4xlarge (16 vCPUs / 32GB)
* c5.9xlarge (36 vCPUs / 72GB)
> *Intel® Xeon® Platinum 8275L*
* c5.12xlarge (48 vCPUs / 96GB)
#### Azure
> *Intel® Xeon® Platinum 8168*
* F16s_v2 (16 vCPUs / 32GB)
* F32s_v2 (32 vCPUs / 64GB)
#### GCP
> *Undisclosed Intel® Xeon® Scalable CPU*
* c2-standard-30 (30 vCPUs, 120GB)