roberta-sst2-lora-ep20-lr0p0003-bs16-2025-06-19-1544

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

• Loss: 0.2777
• Accuracy: 0.9300

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.242	0.1188	500	0.2004	0.9392
0.2121	0.2375	1000	0.2096	0.9358
0.2013	0.3563	1500	0.2346	0.9335
0.2112	0.4751	2000	0.2608	0.9220
0.2051	0.5938	2500	0.2309	0.9335
0.177	0.7126	3000	0.2881	0.9106
0.153	0.8314	3500	0.3287	0.9312
0.1692	0.9501	4000	0.2265	0.9358
0.2143	1.0689	4500	0.2424	0.9369
0.213	1.1876	5000	0.2184	0.9346
0.1993	1.3064	5500	0.2582	0.9369
0.1985	1.4252	6000	0.2100	0.9278
0.207	1.5439	6500	0.1782	0.9369
0.2123	1.6627	7000	0.1972	0.9335
0.2078	1.7815	7500	0.2132	0.9312
0.2061	1.9002	8000	0.2149	0.9312
0.1969	2.0190	8500	0.2167	0.9335
0.1999	2.1378	9000	0.2013	0.9346
0.1811	2.2565	9500	0.2512	0.9369
0.1886	2.3753	10000	0.2098	0.9312
0.1962	2.4941	10500	0.2159	0.9323
0.2053	2.6128	11000	0.1919	0.9289
0.164	2.7316	11500	0.2329	0.9289
0.1638	2.8504	12000	0.2661	0.9335
0.1864	2.9691	12500	0.2160	0.9312
0.1568	3.0879	13000	0.2777	0.9300

Framework versions

• PEFT 0.15.2
• Transformers 4.52.4
• Pytorch 2.1.0+cu118
• Datasets 3.6.0
• Tokenizers 0.21.1