cb38f59feb0ce6dce75f796156a6b18a

This model is a fine-tuned version of albert/albert-xlarge-v1 on the contemmcm/trec dataset. It achieves the following results on the evaluation set:

• Loss: 1.6964
• Data Size: 1.0
• Epoch Runtime: 11.0179
• Accuracy: 0.2771
• F1 Macro: 0.0723

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	2.1166	0	0.8389	0.1396	0.0972
No log	1	170	1.6869	0.0078	1.1825	0.2771	0.0723
No log	2	340	1.7454	0.0156	1.1443	0.1333	0.0392
No log	3	510	1.6760	0.0312	1.3449	0.1833	0.0569
No log	4	680	1.5586	0.0625	1.6906	0.4188	0.2388
0.0989	5	850	1.7572	0.125	2.3016	0.1458	0.0642
0.0989	6	1020	1.7699	0.25	3.4911	0.1792	0.0506
1.7171	7	1190	1.7768	0.5	5.9858	0.1333	0.0392
1.6906	8.0	1360	1.6964	1.0	11.0179	0.2771	0.0723

Framework versions

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