sv3.3-bigbird-roberta-large

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

• Loss: 0.2977
• Mse: 0.2977
• Mae: 0.5234
• Rmse: 0.5456

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 4
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Mae	Rmse
0.1285	0.1548	50	0.4284	0.4284	0.6358	0.6545
0.1019	0.3096	100	0.4425	0.4425	0.6468	0.6652
0.1012	0.4644	150	0.3164	0.3164	0.5409	0.5625
0.0946	0.6192	200	0.4275	0.4275	0.6351	0.6538
0.0911	0.7740	250	0.3013	0.3013	0.5269	0.5489
0.0948	0.9288	300	0.1954	0.1954	0.4166	0.4421
0.1215	1.0836	350	0.4126	0.4126	0.6233	0.6423
0.0872	1.2384	400	0.3116	0.3116	0.5365	0.5582
0.0972	1.3932	450	0.3407	0.3407	0.5629	0.5837
0.0953	1.5480	500	0.3546	0.3546	0.5750	0.5955
0.0913	1.7028	550	0.3799	0.3799	0.5965	0.6164
0.0948	1.8576	600	0.2853	0.2853	0.5115	0.5342
0.0912	2.0124	650	0.3401	0.3401	0.5623	0.5832
0.09	2.1672	700	0.3295	0.3295	0.5528	0.5740
0.0956	2.3220	750	0.3267	0.3267	0.5503	0.5715
0.0944	2.4768	800	0.3099	0.3099	0.5349	0.5567
0.0886	2.6316	850	0.3036	0.3036	0.5290	0.5510
0.0931	2.7864	900	0.3109	0.3109	0.5358	0.5576
0.087	2.9412	950	0.2977	0.2977	0.5234	0.5456

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu126
• Datasets 3.3.2
• Tokenizers 0.21.0