56793e314287979c6fc67d40e2c84baf

This model is a fine-tuned version of albert/albert-large-v2 on the contemmcm/cls_20newsgroups dataset. It achieves the following results on the evaluation set:

• Loss: 3.0018
• Data Size: 1.0
• Epoch Runtime: 80.5163
• Accuracy: 0.0491
• F1 Macro: 0.0047

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	3.0466	0	6.9500	0.0433	0.0140
No log	1	499	3.0461	0.0078	7.8486	0.0411	0.0212
0.0304	2	998	3.0421	0.0156	8.1625	0.0491	0.0047
0.0551	3	1497	3.0290	0.0312	9.3625	0.0504	0.0088
0.1034	4	1996	3.0270	0.0625	11.6978	0.0502	0.0048
3.0327	5	2495	3.0179	0.125	16.2858	0.0544	0.0052
3.027	6	2994	3.0070	0.25	25.4601	0.0532	0.0050
3.0094	7	3493	3.0087	0.5	43.8611	0.0507	0.0048
3.0048	8.0	3992	2.9986	1.0	80.7054	0.0496	0.0047
3.0038	9.0	4491	3.0071	1.0	80.6288	0.0491	0.0047
2.9998	10.0	4990	3.0018	1.0	80.6492	0.0444	0.0042
2.9978	11.0	5489	3.0005	1.0	80.5434	0.0502	0.0048
2.9983	12.0	5988	3.0018	1.0	80.5163	0.0491	0.0047

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1