source-type-model

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

• Loss: 0.6271
• F1: 0.6772

Classifies the following tags:

    'Cannot Determine'
    'Report/Document'
    'Named Individual'
    'Unnamed Individual'
    'Database'
    'Unnamed Group'
    'Named Group'
    'Vote/Poll'

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: 5
• eval_batch_size: 5
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	0.12	100	0.7192	0.3792
No log	0.25	200	0.7716	0.4005
No log	0.37	300	0.7565	0.5297
No log	0.49	400	0.5788	0.5806
0.8223	0.62	500	0.5402	0.5933
0.8223	0.74	600	0.5032	0.6666
0.8223	0.86	700	0.4658	0.6754
0.8223	0.99	800	0.5359	0.6441
0.8223	1.11	900	0.5295	0.6442
0.6009	1.23	1000	0.6077	0.6597
0.6009	1.35	1100	0.6169	0.6360
0.6009	1.48	1200	0.6014	0.6277
0.6009	1.6	1300	0.6382	0.6327
0.6009	1.72	1400	0.5226	0.6787
0.5644	1.85	1500	0.4922	0.6485
0.5644	1.97	1600	0.6181	0.6517
0.5644	2.09	1700	0.6106	0.6781
0.5644	2.22	1800	0.6652	0.6760
0.5644	2.34	1900	0.6252	0.6739
0.3299	2.46	2000	0.6620	0.6606
0.3299	2.59	2100	0.6317	0.6772
0.3299	2.71	2200	0.6170	0.6726
0.3299	2.83	2300	0.6400	0.6773
0.3299	2.96	2400	0.6271	0.6772

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu117
• Datasets 2.13.1
• Tokenizers 0.13.3