Wollaston/gelato-roberta-base

roberta-base

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

• Loss: 0.0722
• Precision: 0.9479
• Recall: 0.9433
• F1: 0.9456
• Accuracy: 0.9887

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: 0.2
• num_epochs: 47

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	10	1.9057	0.0256	0.0032	0.0058	0.7676
No log	2.0	20	0.8889	0.0	0.0	0.0	0.7693
No log	3.0	30	0.5648	0.4330	0.3404	0.3811	0.8368
No log	4.0	40	0.2771	0.7138	0.7358	0.7247	0.9308
No log	5.0	50	0.1407	0.8482	0.8606	0.8544	0.9667
No log	6.0	60	0.0868	0.8369	0.8898	0.8625	0.9795
No log	7.0	70	0.0635	0.8640	0.9060	0.8845	0.9826
No log	8.0	80	0.0513	0.9022	0.9271	0.9145	0.9875
No log	9.0	90	0.0610	0.8947	0.9222	0.9082	0.9855
No log	10.0	100	0.0626	0.9116	0.9190	0.9153	0.9828
No log	11.0	110	0.0792	0.9221	0.9206	0.9213	0.9854
No log	12.0	120	0.0617	0.9019	0.9384	0.9198	0.9840
No log	13.0	130	0.0486	0.8955	0.9303	0.9126	0.9866
No log	14.0	140	0.0615	0.9055	0.9319	0.9185	0.9860
No log	15.0	150	0.0566	0.9401	0.9417	0.9409	0.9889
No log	16.0	160	0.0433	0.9220	0.9384	0.9301	0.9894
No log	17.0	170	0.0591	0.9247	0.9352	0.9299	0.9881
No log	18.0	180	0.0497	0.9311	0.9417	0.9363	0.9891
No log	19.0	190	0.0657	0.9493	0.9400	0.9446	0.9885
No log	20.0	200	0.0673	0.9416	0.9400	0.9408	0.9878
No log	21.0	210	0.0636	0.9265	0.9400	0.9332	0.9869
No log	22.0	220	0.0671	0.9385	0.9400	0.9393	0.9879
No log	23.0	230	0.0623	0.9278	0.9368	0.9323	0.9876
No log	24.0	240	0.0588	0.9403	0.9449	0.9426	0.9889
No log	25.0	250	0.0596	0.9309	0.9384	0.9346	0.9879
No log	26.0	260	0.0623	0.9356	0.9417	0.9386	0.9886
No log	27.0	270	0.0617	0.9296	0.9417	0.9356	0.9886
No log	28.0	280	0.0661	0.9415	0.9384	0.9399	0.9882
No log	29.0	290	0.0614	0.9341	0.9417	0.9379	0.9879
No log	30.0	300	0.0638	0.9448	0.9433	0.9440	0.9884
No log	31.0	310	0.0681	0.9464	0.9449	0.9457	0.9890
No log	32.0	320	0.0704	0.9462	0.9400	0.9431	0.9884
No log	33.0	330	0.0680	0.9495	0.9449	0.9472	0.9892
No log	34.0	340	0.0699	0.9447	0.9417	0.9432	0.9887
No log	35.0	350	0.0718	0.9558	0.9465	0.9511	0.9891
No log	36.0	360	0.0724	0.9542	0.9449	0.9495	0.9892
No log	37.0	370	0.0734	0.9478	0.9417	0.9447	0.9887
No log	38.0	380	0.0732	0.9478	0.9417	0.9447	0.9889
No log	39.0	390	0.0729	0.9478	0.9417	0.9447	0.9890
No log	40.0	400	0.0725	0.9494	0.9433	0.9463	0.9890
No log	41.0	410	0.0728	0.9511	0.9449	0.9480	0.9892
No log	42.0	420	0.0722	0.9511	0.9449	0.9480	0.9892
No log	43.0	430	0.0716	0.9511	0.9449	0.9480	0.9893
No log	44.0	440	0.0723	0.9510	0.9433	0.9471	0.9890
No log	45.0	450	0.0723	0.9479	0.9433	0.9456	0.9890
No log	46.0	460	0.0723	0.9479	0.9433	0.9456	0.9887
No log	47.0	470	0.0722	0.9479	0.9433	0.9456	0.9887

Framework versions

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