README.md

metadata

language: en
license: mit
library_name: pytorch

Plainly Optimized Network

Dataset: BIGBENCH

Trainer Hyperparameters:

• lr = 5e-05
• per_device_batch_size = 8
• gradient_accumulation_steps = 2
• weight_decay = 0.0
• seed = 42

eval_loss	eval_accuracy	epoch
10.477	0.571	1.0
10.386	0.571	2.0
10.400	0.571	3.0
10.386	0.571	4.0
10.386	0.571	5.0
10.411	0.571	6.0
10.389	0.571	7.0
10.386	0.571	8.0
10.385	0.571	9.0
10.391	0.571	10.0
10.384	0.571	11.0
10.383	0.571	12.0
10.383	0.571	13.0
10.469	0.571	14.0
10.384	0.571	15.0
10.383	0.571	16.0