fresh-2-layer-medmcqa-distill-of-fresh-2-layer-gpqa-loop-4

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

• Loss: 0.4186
• Accuracy: 0.5051

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	2.8693	0.2677
No log	2.0	126	2.2777	0.3485
No log	3.0	189	1.0399	0.4141
No log	4.0	252	1.8741	0.4293
No log	5.0	315	1.2779	0.4394
No log	6.0	378	0.7112	0.4646
No log	7.0	441	0.8380	0.4596
1.9226	8.0	504	0.7028	0.4697
1.9226	9.0	567	0.6589	0.4848
1.9226	10.0	630	0.6303	0.4495
1.9226	11.0	693	0.7083	0.4646
1.9226	12.0	756	0.4850	0.4899
1.9226	13.0	819	0.5145	0.4848
1.9226	14.0	882	0.7032	0.4697
1.9226	15.0	945	0.4812	0.4697
0.2279	16.0	1008	0.4186	0.5051
0.2279	17.0	1071	0.3735	0.5
0.2279	18.0	1134	0.3894	0.5051
0.2279	19.0	1197	0.3845	0.5051
0.2279	20.0	1260	0.3925	0.5051

Framework versions

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