ee58d7d245640c79e8fa05b2c2502087

This model is a fine-tuned version of albert/albert-large-v2 on the ccdv/patent-classification [abstract] dataset. It achieves the following results on the evaluation set:

• Loss: 1.9850
• Data Size: 1.0
• Epoch Runtime: 64.5639
• Accuracy: 0.2071
• F1 Macro: 0.0381

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	2.2447	0	4.8796	0.0863	0.0470
No log	1	781	2.4177	0.0078	5.6011	0.0745	0.0182
No log	2	1562	2.0395	0.0156	5.8103	0.1266	0.0538
No log	3	2343	1.9996	0.0312	6.7924	0.2091	0.0431
0.0469	4	3124	1.9839	0.0625	8.6366	0.2218	0.0403
2.0372	5	3905	2.0192	0.125	12.2354	0.2071	0.0381
2.0312	6	4686	1.9889	0.25	19.6739	0.2071	0.0381
1.9999	7	5467	1.9851	0.5	34.3743	0.2071	0.0381
1.9715	8.0	6248	1.9850	1.0	64.5639	0.2071	0.0381

Framework versions

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