deberta-v3-base-sst5

This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0544
• F1: 0.5493

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: 64
• eval_batch_size: 64
• 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: 0.01
• num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	F1
1.1525	0.7463	100	1.2348	0.3353
1.0138	1.4925	200	1.0883	0.4924
0.9136	2.2388	300	1.0727	0.5172
0.8806	2.9851	400	1.1005	0.5205
0.8115	3.7313	500	1.0544	0.5493
0.7483	4.4776	600	1.1093	0.5307
0.696	5.2239	700	1.1064	0.5424
0.7167	5.9701	800	1.1407	0.5293
0.7079	6.7164	900	1.1766	0.5206
0.6593	7.4627	1000	1.2103	0.5257
0.6299	8.2090	1100	1.2031	0.5249
0.675	8.9552	1200	1.2117	0.5198
0.5619	9.7015	1300	1.2219	0.5208

Framework versions

• Transformers 4.49.0
• Pytorch 2.5.1
• Datasets 3.2.0
• Tokenizers 0.21.0