Mistral-SYDNEY

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

• Loss: 0.5562

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 64
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.0454	1.0	44	0.9482
0.8165	2.0	88	0.7148
0.6887	3.0	132	0.6592
0.6282	4.0	176	0.6342
0.6038	5.0	220	0.6053
0.5782	6.0	264	0.6054
0.5601	7.0	308	0.5857
0.5489	8.0	352	0.5795
0.5386	9.0	396	0.5744
0.5362	10.0	440	0.5665
0.5271	11.0	484	0.5617
0.5238	12.0	528	0.5617
0.5204	13.0	572	0.5604
0.5157	14.0	616	0.5614
0.5132	15.0	660	0.5622
0.5137	16.0	704	0.5624
0.5101	17.0	748	0.5580
0.5096	18.0	792	0.5571
0.5071	19.0	836	0.5572
0.5055	20.0	880	0.5591
0.5029	21.0	924	0.5574
0.5051	22.0	968	0.5564
0.5019	23.0	1012	0.5584
0.5025	24.0	1056	0.5574
0.5000	25.0	1100	0.5549
0.4982	26.0	1144	0.5558
0.4990	27.0	1188	0.5570
0.4991	28.0	1232	0.5581
0.4973	29.0	1276	0.5558
0.4967	30.0	1320	0.5599
0.4954	31.0	1364	0.5587
0.4951	32.0	1408	0.5555
0.4935	33.0	1452	0.5565
0.4935	34.0	1496	0.5547
0.4925	35.0	1540	0.5573
0.4933	36.0	1584	0.5563
0.4917	37.0	1628	0.5563
0.4915	38.0	1672	0.5583
0.4904	39.0	1716	0.5559
0.4909	40.0	1760	0.5551
0.4889	41.0	1804	0.5558
0.4891	42.0	1848	0.5554
0.4882	43.0	1892	0.5557
0.4877	44.0	1936	0.5562

Framework versions

• Transformers 5.12.0
• Pytorch 2.12.0+cu130
• Datasets 4.8.5
• Tokenizers 0.22.2