fresh-2-layer-piqa16113-distill-of-fresh-2-layer-gpqa_EVAL_gpqa

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 12.5601
• Accuracy: 0.5202

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0005
• train_batch_size: 32
• eval_batch_size: 32
• seed: 321
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.2	100	17.5725	0.2273
No log	0.4	200	13.7753	0.3939
No log	0.6	300	14.6182	0.3687
No log	0.79	400	13.4422	0.3687
2.8312	0.99	500	12.2908	0.4242
2.8312	1.19	600	13.2865	0.3687
2.8312	1.39	700	12.8823	0.4596
2.8312	1.59	800	12.7374	0.4747
2.8312	1.79	900	14.0988	0.4848
1.0908	1.98	1000	12.8163	0.5
1.0908	2.18	1100	13.4111	0.5051
1.0908	2.38	1200	12.8755	0.4697
1.0908	2.58	1300	12.9406	0.5101
1.0908	2.78	1400	13.3437	0.4747
0.6569	2.98	1500	15.6074	0.4899
0.6569	3.17	1600	12.5601	0.5202
0.6569	3.37	1700	13.7042	0.5051
0.6569	3.57	1800	12.6509	0.4848
0.6569	3.77	1900	12.9618	0.5152
0.4619	3.97	2000	13.1769	0.4697
0.4619	4.17	2100	13.0498	0.4646

Framework versions

• Transformers 4.34.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.14.0