afaji/fresh-2-layer-medmcqa-distill-of-fresh-2-layer-gpqa-loop-2

fresh-2-layer-medmcqa-distill-of-fresh-2-layer-medmcqa-distill-of-fresh-2-layer-gpqa

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

• Loss: 0.9186
• Accuracy: 0.5152

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: 16
• eval_batch_size: 16
• seed: 321
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	63	4.2492	0.2727
No log	2.0	126	3.2851	0.3889
No log	3.0	189	2.1889	0.4444
No log	4.0	252	4.3537	0.4646
No log	5.0	315	1.4476	0.4697
No log	6.0	378	1.1196	0.4646
No log	7.0	441	1.5751	0.4646
2.425	8.0	504	0.9802	0.4343
2.425	9.0	567	2.4061	0.4495
2.425	10.0	630	0.9186	0.5152
2.425	11.0	693	0.9569	0.4848
2.425	12.0	756	0.9649	0.4798
2.425	13.0	819	1.3807	0.4899
2.425	14.0	882	0.6900	0.4899
2.425	15.0	945	0.8787	0.4747
0.3146	16.0	1008	0.7985	0.4949
0.3146	17.0	1071	0.9305	0.4899
0.3146	18.0	1134	0.9062	0.4848
0.3146	19.0	1197	0.8571	0.5051
0.3146	20.0	1260	0.8674	0.5

Framework versions

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