mega-base-wikitext-finetuned-INCIBE

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

• Loss: 1.5169
• Accuracy: 0.3850

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: 2e-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: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	196	1.5359	0.3742
No log	2.0	392	1.5257	0.3753
1.5035	3.0	588	1.5320	0.3719
1.5035	4.0	784	1.5231	0.3715
1.5035	5.0	980	1.5203	0.3745
1.4755	6.0	1176	1.5217	0.3742
1.4755	7.0	1372	1.5301	0.3719
1.4531	8.0	1568	1.5131	0.3805
1.4531	9.0	1764	1.5212	0.3783
1.4531	10.0	1960	1.5173	0.3771
1.4426	11.0	2156	1.5190	0.3809
1.4426	12.0	2352	1.5122	0.3794
1.4238	13.0	2548	1.5129	0.3794
1.4238	14.0	2744	1.5176	0.3783
1.4238	15.0	2940	1.5139	0.3783
1.4161	16.0	3136	1.5235	0.3805
1.4161	17.0	3332	1.5125	0.3846
1.4115	18.0	3528	1.5171	0.3827
1.4115	19.0	3724	1.5112	0.3827
1.4115	20.0	3920	1.5123	0.3816
1.4052	21.0	4116	1.5126	0.3827
1.4052	22.0	4312	1.5170	0.3850
1.4004	23.0	4508	1.5135	0.3805
1.4004	24.0	4704	1.5157	0.3809
1.4004	25.0	4900	1.5160	0.3824

Framework versions

• Transformers 4.32.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3