Wollaston/gelato-xlm-roberta-base

xlm-roberta-base

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0754
• Precision: 0.9493
• Recall: 0.9493
• F1: 0.9493
• Accuracy: 0.9891

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: 8
• 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_ratio: 0.3
• num_epochs: 49

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	10	2.1547	0.0187	0.0455	0.0265	0.5147
No log	2.0	20	1.1487	0.0	0.0	0.0	0.7473
No log	3.0	30	0.8744	0.0	0.0	0.0	0.7588
No log	4.0	40	0.5312	0.3256	0.3427	0.3339	0.8466
No log	5.0	50	0.3368	0.4750	0.4808	0.4778	0.9020
No log	6.0	60	0.1807	0.7834	0.8094	0.7962	0.9524
No log	7.0	70	0.1123	0.8493	0.8969	0.8724	0.9747
No log	8.0	80	0.0740	0.8783	0.9213	0.8993	0.9810
No log	9.0	90	0.0590	0.8971	0.9143	0.9056	0.9861
No log	10.0	100	0.0591	0.9210	0.9371	0.9289	0.9879
No log	11.0	110	0.0517	0.9099	0.9353	0.9224	0.9874
No log	12.0	120	0.0487	0.9201	0.9458	0.9328	0.9895
No log	13.0	130	0.0735	0.8835	0.9283	0.9054	0.9830
No log	14.0	140	0.0632	0.8986	0.9301	0.9141	0.9848
No log	15.0	150	0.0639	0.8764	0.9301	0.9025	0.9832
No log	16.0	160	0.0649	0.9210	0.9371	0.9289	0.9872
No log	17.0	170	0.0627	0.9024	0.9371	0.9194	0.9853
No log	18.0	180	0.0717	0.9177	0.9353	0.9264	0.9871
No log	19.0	190	0.0529	0.9348	0.9528	0.9437	0.9876
No log	20.0	200	0.0547	0.9330	0.9493	0.9411	0.9879
No log	21.0	210	0.0580	0.9426	0.9476	0.9451	0.9890
No log	22.0	220	0.0613	0.9715	0.9528	0.9620	0.9902
No log	23.0	230	0.0586	0.9511	0.9528	0.9520	0.9891
No log	24.0	240	0.0674	0.9543	0.9493	0.9518	0.9892
No log	25.0	250	0.0663	0.9362	0.9493	0.9427	0.9894
No log	26.0	260	0.0638	0.9445	0.9528	0.9487	0.9896
No log	27.0	270	0.0645	0.9482	0.9598	0.9540	0.9907
No log	28.0	280	0.0727	0.9510	0.9510	0.9510	0.9886
No log	29.0	290	0.0756	0.9462	0.9528	0.9495	0.9883
No log	30.0	300	0.0762	0.9378	0.9493	0.9435	0.9884
No log	31.0	310	0.0753	0.9443	0.9476	0.9459	0.9891
No log	32.0	320	0.0755	0.9459	0.9476	0.9467	0.9891
No log	33.0	330	0.0771	0.9443	0.9476	0.9459	0.9892
No log	34.0	340	0.0783	0.9459	0.9476	0.9467	0.9894
No log	35.0	350	0.0783	0.9476	0.9476	0.9476	0.9895
No log	36.0	360	0.0784	0.9476	0.9476	0.9476	0.9896
No log	37.0	370	0.0829	0.9420	0.9371	0.9395	0.9888
No log	38.0	380	0.0816	0.9457	0.9441	0.9449	0.9895
No log	39.0	390	0.0789	0.9443	0.9476	0.9459	0.9895
No log	40.0	400	0.0780	0.9461	0.9510	0.9486	0.9894
No log	41.0	410	0.0832	0.9473	0.9423	0.9448	0.9879
No log	42.0	420	0.0815	0.9457	0.9441	0.9449	0.9888
No log	43.0	430	0.0785	0.9493	0.9493	0.9493	0.9896
No log	44.0	440	0.0758	0.9494	0.9510	0.9502	0.9892
No log	45.0	450	0.0748	0.9460	0.9493	0.9476	0.9890
No log	46.0	460	0.0751	0.9493	0.9493	0.9493	0.9891
No log	47.0	470	0.0755	0.9476	0.9476	0.9476	0.9890
No log	48.0	480	0.0755	0.9493	0.9493	0.9493	0.9891
No log	49.0	490	0.0754	0.9493	0.9493	0.9493	0.9891

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.1.1
• Tokenizers 0.22.1