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: 4.7435
• Bertscore Precision Top1: 69.4231
• Bertscore Recall Top1: 73.3247
• Bertscore F1 Top1: 71.2742
• Bertscore Precision Top1 Mean: 69.4231
• Bertscore Recall Top1 Mean: 73.3247
• Bertscore F1 Top1 Mean: 71.2742
• Bertscore Precision Top3: 72.4567
• Bertscore Recall Top3: 75.8122
• Bertscore F1 Top3: 73.9344
• Bertscore Precision Top3 Mean: 69.3369
• Bertscore Recall Top3 Mean: 73.4405
• Bertscore F1 Top3 Mean: 71.2808
• Bertscore Precision Top5: 73.9557
• Bertscore Recall Top5: 76.7091
• Bertscore F1 Top5: 75.1040
• Bertscore Precision Top5 Mean: 69.3624
• Bertscore Recall Top5 Mean: 73.4929
• Bertscore F1 Top5 Mean: 71.3172
• Bertscore Precision Top10: 75.3599
• Bertscore Recall Top10: 77.8343
• Bertscore F1 Top10: 76.3667
• Bertscore Precision Top10 Mean: 69.1599
• Bertscore Recall Top10 Mean: 73.4324
• Bertscore F1 Top10 Mean: 71.1816

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: 2e-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: 0.06
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Bertscore Precision Top1	Bertscore Recall Top1	Bertscore F1 Top1	Bertscore Precision Top1 Mean	Bertscore Recall Top1 Mean	Bertscore F1 Top1 Mean	Bertscore Precision Top3	Bertscore Recall Top3	Bertscore F1 Top3	Bertscore Precision Top3 Mean	Bertscore Recall Top3 Mean	Bertscore F1 Top3 Mean	Bertscore Precision Top5	Bertscore Recall Top5	Bertscore F1 Top5	Bertscore Precision Top5 Mean	Bertscore Recall Top5 Mean	Bertscore F1 Top5 Mean	Bertscore Precision Top10	Bertscore Recall Top10	Bertscore F1 Top10	Bertscore Precision Top10 Mean	Bertscore Recall Top10 Mean	Bertscore F1 Top10 Mean
No log	1.0	606	4.9901	68.7520	72.9461	70.7337	68.7520	72.9461	70.7337	72.1198	75.4872	73.6212	68.8316	73.0485	70.8266	73.1629	76.2283	74.4758	68.6738	72.9180	70.6819	75.1303	77.6449	76.1507	68.7304	72.9866	70.7462
No log	2.0	1212	4.8858	69.2416	73.3064	71.1667	69.2416	73.3064	71.1667	72.4188	75.6314	73.8351	69.2952	73.4484	71.2634	73.5839	76.4682	74.8175	69.2015	73.3693	71.1773	75.2007	77.7680	76.2653	69.0913	73.3655	71.1144
No log	3.0	1818	4.8845	69.0534	73.1345	70.9860	69.0534	73.1345	70.9860	72.2111	75.4684	73.6579	69.0691	73.3228	71.0816	73.3766	76.4212	74.6816	69.0125	73.3447	71.0603	75.1277	77.7571	76.2173	68.9938	73.3990	71.0762
No log	4.0	2424	4.8869	69.3035	73.2836	71.1877	69.3035	73.2836	71.1877	72.4419	75.6757	73.8648	69.3172	73.4788	71.2876	73.6316	76.5581	74.8813	69.1884	73.4034	71.1839	75.1194	77.7941	76.1902	69.0444	73.3921	71.1014
No log	5.0	3030	4.8059	69.4452	73.4349	71.3335	69.4452	73.4349	71.3335	72.6238	75.8750	74.0494	69.4236	73.5510	71.3804	73.8090	76.7330	75.0386	69.3161	73.4983	71.2958	75.3875	77.8775	76.4057	69.1467	73.4308	71.1751
No log	6.0	3636	4.8282	69.3672	73.3103	71.2332	69.3672	73.3103	71.2332	72.5659	75.7743	73.9856	69.3733	73.4869	71.3226	73.8688	76.7862	75.1133	69.2316	73.4670	71.2346	75.3331	77.9016	76.3904	69.0741	73.3896	71.1164
No log	7.0	4242	4.7246	69.3667	73.3375	71.2496	69.3667	73.3375	71.2496	72.5054	75.7440	73.9124	69.4398	73.5218	71.3728	73.6991	76.6942	74.9626	69.2684	73.4909	71.2654	75.1182	77.8364	76.2124	69.1234	73.4519	71.1719
No log	8.0	4848	4.7093	69.4197	73.3537	71.2865	69.4197	73.3537	71.2865	72.5558	75.8420	73.9847	69.3971	73.5186	71.3496	73.9247	76.7740	75.1221	69.2820	73.4882	71.2719	75.2892	77.8283	76.3209	69.1216	73.4226	71.1562
3.4957	9.0	5454	4.8471	69.4257	73.2362	71.2324	69.4257	73.2362	71.2324	72.3364	75.7030	73.8146	69.3297	73.4323	71.2719	73.9476	76.6215	75.0459	69.3692	73.4637	71.3062	75.4202	77.8834	76.4114	69.2006	73.4434	71.2081
3.4957	10.0	6060	4.6768	69.4231	73.3247	71.2742	69.4231	73.3247	71.2742	72.4567	75.8122	73.9344	69.3369	73.4405	71.2808	73.9557	76.7091	75.1040	69.3624	73.4929	71.3172	75.3599	77.8343	76.3667	69.1599	73.4324	71.1816

Framework versions

• Transformers 5.8.0
• Pytorch 2.11.0+cu126
• Datasets 3.6.0
• Tokenizers 0.22.2