distilbert-base-uncased_08112024T130901

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

• Loss: 0.4617
• F1: 0.8549
• Learning Rate: 0.0

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 600
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Rate
No log	0.9942	86	1.7918	0.1082	0.0000
No log	2.0	173	1.7405	0.2971	0.0000
No log	2.9942	259	1.4964	0.4840	0.0000
No log	4.0	346	1.2746	0.5281	0.0000
No log	4.9942	432	1.1099	0.5611	0.0000
1.4707	6.0	519	0.9946	0.5913	0.0000
1.4707	6.9942	605	0.8987	0.6406	1e-05
1.4707	8.0	692	0.8116	0.6910	0.0000
1.4707	8.9942	778	0.7320	0.7265	0.0000
1.4707	10.0	865	0.6655	0.7655	0.0000
1.4707	10.9942	951	0.6296	0.7802	0.0000
0.7266	12.0	1038	0.5789	0.8028	0.0000
0.7266	12.9942	1124	0.5481	0.8202	0.0000
0.7266	14.0	1211	0.4914	0.8420	0.0000
0.7266	14.9942	1297	0.4787	0.8476	0.0000
0.7266	16.0	1384	0.4776	0.8510	0.0000
0.7266	16.9942	1470	0.4633	0.8554	0.0000
0.3466	18.0	1557	0.4627	0.8547	5e-07
0.3466	18.9942	1643	0.4617	0.8549	1e-07
0.3466	19.8844	1720	0.4621	0.8546	0.0

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.19.1