db_fe2_17.1.4

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3102
• Accuracy: 0.9535
• Balanced Accuracy: 0.9593
• F1 Weighted: 0.9531
• Precision Weighted: 0.9532
• Recall Weighted: 0.9535
• F1 Macro: 0.9574
• Precision Macro: 0.9560
• Recall Macro: 0.9593
• F1 Min: 0.7649
• N Below 80: 1
• N Errors: 514
• Conf Err Rate: 0.4650

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: 3.5e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 14
• label_smoothing_factor: 0.01

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Balanced Accuracy	F1 Weighted	Precision Weighted	Recall Weighted	F1 Macro	Precision Macro	Recall Macro	F1 Min	N Below 80	N Errors	Conf Err Rate
1.2937	1.0	978	0.8620	0.8140	0.8240	0.8102	0.8146	0.8140	0.8185	0.8213	0.8240	0.4906	27	2054	0.0073
0.4703	2.0	1956	0.4143	0.9097	0.9202	0.9077	0.9111	0.9097	0.9156	0.9149	0.9202	0.5986	6	997	0.1464
0.2915	3.0	2934	0.3348	0.9343	0.9415	0.9344	0.9355	0.9343	0.9406	0.9409	0.9415	0.6977	1	725	0.2152
0.2153	4.0	3912	0.3283	0.9387	0.9469	0.9382	0.9391	0.9387	0.9439	0.9422	0.9469	0.7074	2	677	0.2541
0.1707	5.0	4890	0.3223	0.9449	0.9512	0.9450	0.9460	0.9449	0.9502	0.9501	0.9512	0.7627	1	608	0.3290
0.1416	6.0	5868	0.3264	0.9451	0.9511	0.9449	0.9458	0.9451	0.9498	0.9495	0.9511	0.7543	2	606	0.3878
0.1241	7.0	6846	0.3290	0.9466	0.9542	0.9457	0.9469	0.9466	0.9509	0.9490	0.9542	0.7157	2	590	0.4220
0.1138	8.0	7824	0.3177	0.9509	0.9581	0.9503	0.9512	0.9509	0.9555	0.9540	0.9581	0.7341	1	542	0.4428
0.1094	9.0	8802	0.3195	0.9512	0.9578	0.9507	0.9511	0.9512	0.9554	0.9539	0.9578	0.7529	1	539	0.4675
0.1065	10.0	9780	0.3119	0.9528	0.9593	0.9525	0.9528	0.9528	0.9571	0.9555	0.9593	0.7575	1	521	0.4453
0.1034	11.0	10758	0.3099	0.9532	0.9595	0.9528	0.9530	0.9532	0.9573	0.9557	0.9595	0.7565	1	517	0.4739
0.1018	12.0	11736	0.3099	0.9530	0.9592	0.9526	0.9529	0.9530	0.9572	0.9558	0.9592	0.7607	1	519	0.4682
0.1024	13.0	12714	0.3100	0.9535	0.9593	0.9532	0.9533	0.9535	0.9575	0.9561	0.9593	0.7621	1	513	0.4717
0.1011	14.0	13692	0.3102	0.9535	0.9593	0.9531	0.9532	0.9535	0.9574	0.9560	0.9593	0.7649	1	514	0.4650

Framework versions

• Transformers 5.16.1
• Pytorch 2.11.0+cu128
• Datasets 4.0.0
• Tokenizers 0.23.1