gpt2_sm_cv_defined_4

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

• Loss: 2.4859
• Accuracy: 0.777
• Precision: 0.3889
• Recall: 0.2513
• F1: 0.3053
• D-index: 1.4874

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 8000
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	D-index
No log	1.0	250	1.2496	0.469	0.1830	0.4974	0.2676	1.1202
2.4045	2.0	500	0.6575	0.679	0.2056	0.2256	0.2152	1.3395
2.4045	3.0	750	0.5358	0.784	0.2439	0.0513	0.0847	1.4262
0.5054	4.0	1000	0.5534	0.786	0.3333	0.0974	0.1508	1.4457
0.5054	5.0	1250	0.5221	0.785	0.2727	0.0615	0.1004	1.4313
0.4682	6.0	1500	0.5396	0.765	0.3276	0.1949	0.2444	1.4510
0.4682	7.0	1750	0.5442	0.766	0.3415	0.2154	0.2642	1.4596
0.4097	8.0	2000	0.7828	0.79	0.2581	0.0410	0.0708	1.4309
0.4097	9.0	2250	0.6443	0.771	0.3607	0.2256	0.2776	1.4701
0.3341	10.0	2500	0.6839	0.76	0.3529	0.2769	0.3103	1.4727
0.3341	11.0	2750	0.7968	0.725	0.3095	0.3333	0.3210	1.4434
0.2456	12.0	3000	1.0615	0.771	0.3534	0.2103	0.2637	1.4648
0.2456	13.0	3250	1.7036	0.786	0.3797	0.1538	0.2190	1.4657
0.1537	14.0	3500	1.5848	0.741	0.3280	0.3128	0.3202	1.4587
0.1537	15.0	3750	1.5904	0.727	0.3125	0.3333	0.3226	1.4462
0.1323	16.0	4000	1.9229	0.685	0.3	0.4615	0.3636	1.4311
0.1323	17.0	4250	2.2383	0.785	0.3684	0.1436	0.2066	1.4607
0.1117	18.0	4500	2.4084	0.799	0.4167	0.0769	0.1299	1.4564
0.1117	19.0	4750	2.5225	0.798	0.4426	0.1385	0.2109	1.4769
0.0759	20.0	5000	2.4859	0.777	0.3889	0.2513	0.3053	1.4874

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3