https://github.com/enijkamp/metrics_generative
metrics for generative models (fid, inception)
https://github.com/enijkamp/metrics_generative
generative-model inception-v3 pytorch tensorflow
Last synced: about 1 year ago
JSON representation
metrics for generative models (fid, inception)
- Host: GitHub
- URL: https://github.com/enijkamp/metrics_generative
- Owner: enijkamp
- Created: 2018-09-13T19:04:35.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2018-09-14T03:42:28.000Z (almost 8 years ago)
- Last Synced: 2025-04-30T03:39:30.672Z (about 1 year ago)
- Topics: generative-model, inception-v3, pytorch, tensorflow
- Language: Python
- Homepage:
- Size: 17.6 KB
- Stars: 13
- Watchers: 3
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# metrics for generative models
Updated inception and fid scores code.
requires:
* python >= 3.6
* pytorch >= 0.4.0
* tensorflow >= 1.10
features:
* fixed open-ai tensorflow inception v2 code with bs=100 for fast computation (change bs=100 -> 1 is slow, see https://github.com/openai/improved-gan/commit/0b7ed92e47ff7047701be3e10a3bd6363999f5e7)
* tests for both fid and inception v2 and v3 scores
* invoking tensorflow inception v2 score from within pytorch session
notes:
* be careful when computing inception scores, verify (1) nhwc, ncwh, rgb order, (2) normalization, (3) sample size
* see general inception score issues: https://github.com/sbarratt/inception-score-pytorch/issues?q=is%3Aissue+is%3Aclosed
* the official, widely used open-ai tensorflow v2 code appears to have a bug: https://github.com/sbarratt/inception-score-pytorch/issues/1
* read about general issues regarding the inception score: https://arxiv.org/pdf/1801.01973.pdf
* do NOT use the inception score for datasets other than imagenet (even though in the code we used it for cifar), use fid score instead