gpt2-finetuned-mcqa

This model is a fine-tuned version of openai-community/gpt2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.4696
• Bertscore Precision: 0.1390
• Bertscore Recall: 0.1471
• Bertscore F1: 0.1428

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Bertscore Precision	Bertscore Recall	Bertscore F1
4.6412	1.0000	11719	2.9435	0.1223	0.1435	0.1316
4.4659	2.0	23439	2.8198	0.1268	0.1445	0.1346
4.1443	3.0000	35158	2.7298	0.1326	0.1452	0.1384
4.1037	4.0	46878	2.6608	0.1348	0.1457	0.1399
4.0107	5.0000	58597	2.5980	0.1338	0.1462	0.1395
4.0216	6.0	70317	2.5494	0.1353	0.1463	0.1404
3.5677	7.0000	82036	2.5156	0.1371	0.1466	0.1415
3.619	8.0	93756	2.4859	0.1376	0.1468	0.1419
3.5389	9.0000	105475	2.4735	0.1391	0.1470	0.1429
3.713	9.9996	117190	2.4696	0.1390	0.1471	0.1428

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1