966071a1ef4a83ba5529f0e409c2d5c3

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

• Loss: 1.6998
• Data Size: 1.0
• Epoch Runtime: 11.6907
• Accuracy: 0.1792
• F1 Macro: 0.0506

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	2.1425	0	0.9167	0.1333	0.0392
No log	1	170	1.8314	0.0078	1.3491	0.1333	0.0392
No log	2	340	1.7703	0.0156	1.1206	0.1333	0.0392
No log	3	510	1.8289	0.0312	1.3793	0.1958	0.1090
No log	4	680	1.7402	0.0625	1.6935	0.1833	0.0742
0.1045	5	850	1.7204	0.125	2.3872	0.1375	0.0556
0.1045	6	1020	1.7374	0.25	3.7592	0.1792	0.0506
1.682	7	1190	1.7369	0.5	6.3381	0.1333	0.0392
1.679	8.0	1360	1.6653	1.0	11.8504	0.2771	0.0723
1.68	9.0	1530	1.7156	1.0	11.7201	0.1646	0.0471
1.6813	10.0	1700	1.6713	1.0	11.7804	0.2771	0.0723
1.688	11.0	1870	1.6825	1.0	11.5884	0.1792	0.0506
1.6645	12.0	2040	1.6998	1.0	11.6907	0.1792	0.0506

Framework versions

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