gs-Logion

This model is a fine-tuned version of cabrooks/LOGION-50k_wordpiece on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.6266
• Top1: 30.0
• Top5: 40.0
• Top10: 44.6667
• Top20: 46.3333
• Bertscore F1 Top1: 84.3060
• Bertscore F1 Top1 Mean: 84.3060
• Bertscore F1 Top5: 88.2838
• Bertscore F1 Top5 Mean: 79.8681
• Bertscore F1 Top10: 89.5413
• Bertscore F1 Top10 Mean: 78.3595
• Bertscore F1 Top20: 90.5304
• Bertscore F1 Top20 Mean: 77.3756
• Cos Sim Top1 Max: 63.6562
• Cos Sim Top1 Mean: 63.6562
• Cos Sim Top5 Max: 73.5479
• Cos Sim Top5 Mean: 53.6057
• Cos Sim Top10 Max: 77.6688
• Cos Sim Top10 Mean: 51.5711
• Cos Sim Top20 Max: 80.1120
• Cos Sim Top20 Mean: 50.2923
• Composite Score: 46.8281

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: 2.107981566771331e-05
• train_batch_size: 128
• 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
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Top1	Top5	Top10	Top20	Bertscore F1 Top1	Bertscore F1 Top1 Mean	Bertscore F1 Top5	Bertscore F1 Top5 Mean	Bertscore F1 Top10	Bertscore F1 Top10 Mean	Bertscore F1 Top20	Bertscore F1 Top20 Mean	Cos Sim Top1 Max	Cos Sim Top1 Mean	Cos Sim Top5 Max	Cos Sim Top5 Mean	Cos Sim Top10 Max	Cos Sim Top10 Mean	Cos Sim Top20 Max	Cos Sim Top20 Mean	Composite Score
2.9119	1.0	136	2.6962	27.6667	39.0	43.0	44.6667	83.7187	83.7187	88.1704	79.8185	89.3254	78.3467	90.4031	77.3696	62.4214	62.4214	73.3831	53.7380	76.9649	51.5859	79.7153	50.4135	45.0440
2.8573	2.0	272	2.6577	30.0	39.6667	44.3333	46.0	84.3331	84.3331	88.2379	79.8180	89.4981	78.3390	90.4987	77.3429	63.6594	63.6594	73.4449	53.5570	77.4718	51.4226	80.0297	50.2342	46.8297
2.7978	3.0	408	2.6293	30.0	40.0	44.6667	46.3333	84.3060	84.3060	88.2838	79.8681	89.5413	78.3595	90.5304	77.3756	63.6562	63.6562	73.5479	53.6057	77.6688	51.5711	80.1120	50.2923	46.8281

Framework versions

• Transformers 5.9.0
• Pytorch 2.12.0+cu126
• Datasets 3.6.0
• Tokenizers 0.22.2