bert-large-cased-sigir-support-refute-no-label-40

This model is a fine-tuned version of bert-large-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8371

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: 4e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 40.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.4511	1.0	252	2.0790
2.0373	2.0	504	1.8538
1.8052	3.0	756	1.6633
1.6663	4.0	1008	1.5591
1.5556	5.0	1260	1.4441
1.4505	6.0	1512	1.3836
1.3619	7.0	1764	1.3255
1.2968	8.0	2016	1.2505
1.2332	9.0	2268	1.2165
1.1788	10.0	2520	1.1517
1.1408	11.0	2772	1.1446
1.0992	12.0	3024	1.1512
1.0578	13.0	3276	1.1058
1.0277	14.0	3528	1.0662
1.0036	15.0	3780	1.0270
0.9655	16.0	4032	1.0207
0.9364	17.0	4284	1.0220
0.9085	18.0	4536	0.9874
0.8897	19.0	4788	0.9658
0.8661	20.0	5040	0.9603
0.8434	21.0	5292	0.9754
0.8248	22.0	5544	0.9406
0.8052	23.0	5796	0.9154
0.7975	24.0	6048	0.8760
0.7854	25.0	6300	0.8688
0.7673	26.0	6552	0.8536
0.7463	27.0	6804	0.8544
0.7412	28.0	7056	0.8514
0.7319	29.0	7308	0.8356
0.7143	30.0	7560	0.8832
0.7081	31.0	7812	0.8421
0.7026	32.0	8064	0.8295
0.687	33.0	8316	0.8401
0.6882	34.0	8568	0.8053
0.679	35.0	8820	0.8438
0.6672	36.0	9072	0.8450
0.6669	37.0	9324	0.8231
0.6665	38.0	9576	0.8410
0.6596	39.0	9828	0.7909
0.6556	40.0	10080	0.8019

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2