Baseline | moritzm00 Overview – Weights & Biases

Skip to main content

Moritzm00's group workspace

Group: Baseline

1-9

of 9

Tags

Notes

Baseline CNN Model with image augmentation (tf.data API)

The image augmentation with tf.data is much faster, leading to an improvement from over 200s per epoch to just 40s per epoch (this run)

Tags

Baseline

Image Augmentation

tf.data API

Author

State

Finished

Start time

April 23rd, 2023 12:12:48 PM

Runtime

18m 56s

Tracked hours

18m 44s

Run path

moritzm00/UTKFace-Age-Regression/nzdf3wdt

OS

Linux-5.15.90+-x86_64-with-debian-bullseye-sid

Python version

3.7.12

Command

kaggle.ipynb

System Hardware

CPU count	1
Logical CPU count	2
GPU count	1
GPU type	Tesla P100-PCIE-16GB

W&B CLI Version

0.14.0

Group

Config parameters are your model's inputs. Learn more

▶
Config parameters:{} 14 keys
- architecture:
  "Simple CNN"
- batch_size:
  32
- decay_rate:
  0.96
- decay_steps:
  100,000
- early_stopping_mode:
  "min"
- early_stopping_monitor:
  "val_mae"
- early_stopping_patience:
  5
- epochs:
  50
- initial_learning_rate:
  0.001
- loss:
  "mean_absolute_error"
- lr_schedule:
  "ExponentialDecay"
- optimizer:
  "Adam"
- random_rotation:
  0.15
- random_translation:
  0.1

Summary metrics are your model's outputs. Learn more

▶
Summary metrics:{} 6 keys
- epoch/epoch:
  21
- epoch/learning_rate:
  0.0009946889476850629
- epoch/loss:
  8.640949249267578
- epoch/mae:
  8.640949249267578
- epoch/val_loss:
  8.581686973571777
- epoch/val_mae:
  8.581686973571777

This run consumed these artifacts as inputs. Learn more

Loading...

This run produced these artifacts as outputs. Total: 45. Learn more