11dd5bce6aefed726325d637d7914c10

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

• Loss: 0.4248
• Data Size: 1.0
• Epoch Runtime: 48.9494
• Accuracy: 0.8926
• F1 Macro: 0.8923

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	2.9995	0	3.7480	0.0590	0.0213
No log	1	499	3.0035	0.0078	4.4587	0.0837	0.0283
0.0296	2	998	2.7713	0.0156	4.8309	0.2228	0.1684
0.0522	3	1497	2.0281	0.0312	6.0604	0.5597	0.5183
0.0807	4	1996	1.2187	0.0625	7.7718	0.6721	0.6565
1.0848	5	2495	0.8341	0.125	10.6203	0.7490	0.7333
0.6901	6	2994	0.6131	0.25	16.5225	0.7959	0.7843
0.5116	7	3493	0.4907	0.5	27.4050	0.8432	0.8427
0.3597	8.0	3992	0.3928	1.0	49.7849	0.8732	0.8722
0.2877	9.0	4491	0.4060	1.0	49.7398	0.8712	0.8696
0.1741	10.0	4990	0.4009	1.0	50.4551	0.8856	0.8840
0.1474	11.0	5489	0.4299	1.0	48.6970	0.8884	0.8873
0.1441	12.0	5988	0.4248	1.0	48.9494	0.8926	0.8923

Framework versions

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