mitre-gpt2-base

This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.3404

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss
3.2933	2.72	1000	2.6028
2.2619	5.45	2000	2.4654
1.9152	8.17	3000	2.3952
1.6434	10.9	4000	2.3729
1.4289	13.62	5000	2.4208
1.2627	16.35	6000	2.4845
1.1301	19.07	7000	2.5619
1.0169	21.8	8000	2.6058
0.93	24.52	9000	2.6773
0.8587	27.25	10000	2.7389
0.8032	29.97	11000	2.7639
0.7506	32.7	12000	2.8329
0.7079	35.42	13000	2.8934
0.6781	38.15	14000	2.9175
0.6461	40.87	15000	2.9532
0.6205	43.6	16000	3.0008
0.5987	46.32	17000	3.0539
0.5811	49.05	18000	3.0738
0.564	51.77	19000	3.0972
0.5491	54.5	20000	3.1341
0.5377	57.22	21000	3.1558
0.5255	59.95	22000	3.1723
0.516	62.67	23000	3.1984
0.5077	65.4	24000	3.2163
0.5021	68.12	25000	3.2396
0.4946	70.84	26000	3.2413
0.4871	73.57	27000	3.2708
0.4845	76.29	28000	3.2833
0.4791	79.02	29000	3.2847
0.4739	81.74	30000	3.2950
0.4704	84.47	31000	3.3124
0.4678	87.19	32000	3.3122
0.4642	89.92	33000	3.3260
0.4617	92.64	34000	3.3326
0.4605	95.37	35000	3.3325
0.4576	98.09	36000	3.3404

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2