Skip to main content
stacey
Projects
pytorch_intro
Reports
RNN

RNN

new experiments 11/19
Stacey Svetlichnaya
Created on November 20|Last edited on November 20

RNN Notes

Explore correlation between hidden size and accuracy for a bidirectional RNN: larger hidden size leads to higher acc.

accuracy ∝\propto hidden size




50100150200250300350400450500550batch_size00learning_rate0.400.350.300.250.200.150.100.050.00loss020406080100120140160180200hidden_size98.098.198.298.398.498.598.698.798.898.999.099.1acc
Showing first 10 runs
2k4k6k8kStep93949596979899
fiery-snow-1
vibrant-sun-2
comic-fire-3
eager-oath-4
copper-pyramid-5
noble-bird-6
breezy-energy-7
rich-grass-8
polished-donkey-9
logical-monkey-10
Run set 1
17


Grouped Runs




Run set 1
17
Hidden size
315
Batch size
315


Created with ❤️ on Weights & Biases.

https://wandb.ai/stacey/pytorch_intro/reports/RNN--VmlldzozMjk3ODk
Made with Weights & Biases. Sign up or log in to create reports like this one.