ModernBERT_wine_quality_reviews_ft

This model is a fine-tuned version of answerdotai/ModernBERT-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4820
• Accuracy: 0.7864
• F1: 0.6767

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: 8e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.98) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.7593	0.0497	250	0.5938	0.7402	0.5293
0.6378	0.0993	500	0.6212	0.7308	0.5121
0.6001	0.1490	750	0.5726	0.7499	0.5482
0.5945	0.1987	1000	0.5661	0.7504	0.5390
0.5806	0.2484	1250	0.5519	0.7513	0.5693
0.6042	0.2980	1500	0.5528	0.7535	0.5945
0.5719	0.3477	1750	0.5290	0.7705	0.6045
0.5699	0.3974	2000	0.5247	0.7637	0.5958
0.5622	0.4470	2250	0.5318	0.7700	0.5960
0.5597	0.4967	2500	0.5344	0.7588	0.6152
0.5511	0.5464	2750	0.5767	0.7373	0.5459
0.5464	0.5961	3000	0.5078	0.7787	0.6460
0.5375	0.6457	3250	0.5029	0.7821	0.6175
0.5468	0.6954	3500	0.5027	0.7835	0.6312
0.5232	0.7451	3750	0.5053	0.7765	0.6365
0.5459	0.7948	4000	0.4983	0.7889	0.6723
0.5241	0.8444	4250	0.5092	0.7826	0.6751
0.5094	0.8941	4500	0.5215	0.7831	0.6233
0.5223	0.9438	4750	0.4986	0.7881	0.6243
0.5161	0.9934	5000	0.5013	0.7869	0.6205
0.4474	1.0431	5250	0.5010	0.7847	0.6689
0.4417	1.0928	5500	0.4923	0.7898	0.6251
0.4316	1.1425	5750	0.4968	0.7903	0.6950
0.4244	1.1921	6000	0.4958	0.7867	0.6950
0.4479	1.2418	6250	0.4952	0.7874	0.6864
0.4332	1.2915	6500	0.4957	0.7801	0.6741
0.4357	1.3411	6750	0.4887	0.7864	0.6397
0.439	1.3908	7000	0.5425	0.7654	0.6739
0.4298	1.4405	7250	0.4830	0.7864	0.6849
0.4345	1.4902	7500	0.4970	0.7821	0.6324
0.432	1.5398	7750	0.4890	0.7874	0.6772
0.433	1.5895	8000	0.4920	0.7893	0.6716
0.4364	1.6392	8250	0.5033	0.7818	0.6840
0.4196	1.6889	8500	0.4845	0.7886	0.6674
0.4199	1.7385	8750	0.5046	0.7814	0.6377
0.4121	1.7882	9000	0.5066	0.7845	0.6852
0.4222	1.8379	9250	0.4951	0.7852	0.6669
0.4217	1.8875	9500	0.4820	0.7864	0.6767
0.3973	1.9372	9750	0.4964	0.7872	0.7037
0.4297	1.9869	10000	0.4872	0.7840	0.6811
0.3039	2.0366	10250	0.6112	0.7755	0.6741
0.2271	2.0862	10500	0.6606	0.7797	0.6878
0.2149	2.1359	10750	0.6955	0.7736	0.6823
0.2202	2.1856	11000	0.7086	0.7772	0.6681
0.2226	2.2352	11250	0.6691	0.7743	0.6797
0.2162	2.2849	11500	0.6852	0.7741	0.6702
0.198	2.3346	11750	0.7187	0.7763	0.6592
0.2053	2.3843	12000	0.6847	0.7782	0.6816
0.2099	2.4339	12250	0.7302	0.7748	0.6789
0.214	2.4836	12500	0.7198	0.7726	0.6831
0.1938	2.5333	12750	0.7529	0.7741	0.6390
0.1959	2.5830	13000	0.7467	0.7724	0.6783
0.2012	2.6326	13250	0.7245	0.7736	0.6456
0.1915	2.6823	13500	0.7603	0.7678	0.6823
0.2	2.7320	13750	0.7175	0.7748	0.6795
0.1921	2.7816	14000	0.7458	0.7755	0.6809
0.1895	2.8313	14250	0.7607	0.7729	0.6603
0.1855	2.8810	14500	0.7625	0.7736	0.6705
0.1878	2.9307	14750	0.7598	0.7709	0.6707
0.1899	2.9803	15000	0.7604	0.7729	0.6714

Framework versions

• Transformers 4.48.0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0