fresh-8-layer-medmcqa-distill-of-fresh-8-layer-gpqa

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

• Loss: 17.1123
• Accuracy: 0.5455

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	22.9570	0.2273
No log	2.0	126	21.2108	0.3636
No log	3.0	189	26.1433	0.4040
No log	4.0	252	24.1795	0.3838
No log	5.0	315	17.9657	0.4747
No log	6.0	378	20.0576	0.5354
No log	7.0	441	17.5133	0.5
10.1769	8.0	504	22.3248	0.5101
10.1769	9.0	567	20.7352	0.4848
10.1769	10.0	630	22.9071	0.4596
10.1769	11.0	693	17.8100	0.4899
10.1769	12.0	756	17.9827	0.5202
10.1769	13.0	819	19.2382	0.5
10.1769	14.0	882	18.8849	0.4949
10.1769	15.0	945	17.6397	0.5202
2.2143	16.0	1008	19.0081	0.5101
2.2143	17.0	1071	17.8718	0.5152
2.2143	18.0	1134	17.5239	0.5303
2.2143	19.0	1197	17.1123	0.5455
2.2143	20.0	1260	17.7756	0.5404

Framework versions

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