distilbert-base-uncased-finetuned-sst-2-english_07112024T180048

This model is a fine-tuned version of distilbert/distilbert-base-uncased-finetuned-sst-2-english on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4594
• F1: 0.8486
• 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	1.7769	0.0796	0.0000
No log	2.0	173	1.7161	0.2057	0.0000
No log	2.9942	259	1.6015	0.3386	0.0000
No log	4.0	346	1.4321	0.4359	0.0000
No log	4.9942	432	1.2548	0.5262	0.0000
1.5627	6.0	519	1.0659	0.5697	0.0000
1.5627	6.9942	605	0.9261	0.6140	1e-05
1.5627	8.0	692	0.8197	0.6649	0.0000
1.5627	8.9942	778	0.7274	0.7060	0.0000
1.5627	10.0	865	0.6496	0.7548	0.0000
1.5627	10.9942	951	0.5997	0.7741	0.0000
0.7439	12.0	1038	0.5462	0.8076	0.0000
0.7439	12.9942	1124	0.5300	0.8141	0.0000
0.7439	14.0	1211	0.4898	0.8314	0.0000
0.7439	14.9942	1297	0.4791	0.8382	0.0000
0.7439	16.0	1384	0.4756	0.8399	0.0000
0.7439	16.9942	1470	0.4668	0.8457	0.0000
0.3573	18.0	1557	0.4600	0.8489	5e-07
0.3573	18.9942	1643	0.4594	0.8486	1e-07
0.3573	19.8844	1720	0.4596	0.8483	0.0

Framework versions

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