loose_balanced_cf_seed-63_1e-3

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.1956
• Accuracy: 0.4006

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: 0.001
• train_batch_size: 32
• eval_batch_size: 64
• seed: 63
• gradient_accumulation_steps: 8
• total_train_batch_size: 256
• 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
• lr_scheduler_warmup_steps: 32000
• num_epochs: 20.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
5.9736	0.9994	1485	4.4156	0.2928
4.3066	1.9995	2971	3.9060	0.3332
3.6797	2.9996	4457	3.6355	0.3566
3.4925	3.9997	5943	3.4738	0.3709
3.2645	4.9997	7429	3.3761	0.3799
3.1915	5.9998	8915	3.3184	0.3856
3.0832	6.9999	10401	3.2787	0.3894
3.0425	8.0	11887	3.2506	0.3924
2.9847	8.9994	13372	3.2324	0.3942
2.9552	9.9995	14858	3.2234	0.3955
2.9204	10.9996	16344	3.2134	0.3966
2.8999	11.9997	17830	3.2045	0.3980
2.8789	12.9997	19316	3.2006	0.3985
2.8601	13.9998	20802	3.1995	0.3984
2.8494	14.9999	22288	3.1956	0.3990
2.8313	16.0	23774	3.1950	0.3997
2.8307	16.9994	25259	3.1953	0.3998
2.81	17.9995	26745	3.1945	0.4002
2.8189	18.9996	28231	3.1877	0.4007
2.7972	19.9882	29700	3.1956	0.4006

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.20.0