distilbert-base-uncased-pytorch-issues-classifier

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: 0.1250
• F1 Micro: 0.6960
• F1 Macro: 0.2629
• F1 Weighted: 0.6287
• F1 Samples: 0.6492
• Precision Micro: 0.7690
• Precision Macro: 0.2652
• Recall Micro: 0.6356
• Recall Macro: 0.2619
• Subset Accuracy: 0.1883
• Hamming Loss: 0.0463

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Micro	F1 Macro	F1 Weighted	F1 Samples	Precision Micro	Precision Macro	Recall Micro	Recall Macro	Subset Accuracy	Hamming Loss
0.2157	0.5203	500	0.2007	0.3763	0.0643	0.2856	0.3039	0.6629	0.0942	0.2627	0.0583	0.0260	0.0726
0.1622	1.0406	1000	0.1538	0.5910	0.1681	0.5137	0.5324	0.7259	0.2445	0.4983	0.1575	0.0780	0.0575
0.1376	1.5609	1500	0.1372	0.6663	0.2453	0.6019	0.6204	0.7417	0.2602	0.6049	0.2393	0.1691	0.0505
0.1316	2.0812	2000	0.1288	0.6877	0.2599	0.6209	0.6388	0.7568	0.2625	0.6301	0.2589	0.1811	0.0477
0.1226	2.6015	2500	0.1250	0.6960	0.2629	0.6287	0.6492	0.7690	0.2652	0.6356	0.2619	0.1883	0.0463

Framework versions

• Transformers 4.44.0
• Pytorch 2.10.0+cu128
• Datasets 2.18.0
• Tokenizers 0.19.1