b1537a23b1d810eacd1bda32958f7176

This model is a fine-tuned version of distilbert/distilbert-base-uncased on the contemmcm/cls_20newsgroups dataset. It achieves the following results on the evaluation set:

• Loss: 0.4904
• Data Size: 1.0
• Epoch Runtime: 16.2744
• Accuracy: 0.8871
• F1 Macro: 0.8876

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro
No log	0	0	2.9987	0	1.7341	0.0350	0.0109
No log	1	499	2.9842	0.0078	2.2064	0.1046	0.0270
0.03	2	998	2.8631	0.0156	2.0465	0.2823	0.1910
0.0529	3	1497	2.3667	0.0312	2.3256	0.4657	0.4177
0.0865	4	1996	1.5771	0.0625	2.8698	0.6139	0.5679
1.2925	5	2495	0.9627	0.125	3.7625	0.7324	0.7101
0.6942	6	2994	0.6104	0.25	5.5868	0.8057	0.7987
0.4802	7	3493	0.4634	0.5	9.1897	0.8516	0.8509
0.3268	8.0	3992	0.3762	1.0	16.7458	0.8856	0.8849
0.2591	9.0	4491	0.4257	1.0	16.1054	0.8730	0.8707
0.1484	10.0	4990	0.4363	1.0	16.1890	0.8821	0.8817
0.1476	11.0	5489	0.4618	1.0	16.1673	0.8853	0.8850
0.1128	12.0	5988	0.4904	1.0	16.2744	0.8871	0.8876

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1