roberta-sst2-fine-tuning-shuttleworth-t2

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3491
• Accuracy: 0.9335

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch 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: 0.06
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6887	0.0594	250	0.6925	0.5092
0.3907	0.1188	500	0.2747	0.9048
0.2923	0.1781	750	0.2581	0.8968
0.29	0.2375	1000	0.2650	0.9140
0.2636	0.2969	1250	0.2180	0.9255
0.2501	0.3563	1500	0.2401	0.9278
0.2485	0.4157	1750	0.2941	0.9060
0.2465	0.4751	2000	0.2908	0.9117
0.2255	0.5344	2250	0.3400	0.9094
0.2563	0.5938	2500	0.2383	0.9255
0.239	0.6532	2750	0.2242	0.9300
0.2412	0.7126	3000	0.2647	0.9243
0.2303	0.7720	3250	0.2391	0.9323
0.2479	0.8314	3500	0.2478	0.9312
0.2365	0.8907	3750	0.2826	0.9278
0.2034	0.9501	4000	0.2374	0.9346
0.1981	1.0095	4250	0.2414	0.9289
0.2038	1.0689	4500	0.2832	0.9358
0.1478	1.1283	4750	0.2674	0.9335
0.2093	1.1876	5000	0.2612	0.9392
0.1984	1.2470	5250	0.2418	0.9427
0.1836	1.3064	5500	0.2635	0.9381
0.1827	1.3658	5750	0.2924	0.9243
0.1835	1.4252	6000	0.2754	0.9289
0.1911	1.4846	6250	0.2762	0.9346
0.1874	1.5439	6500	0.2190	0.9381
0.1918	1.6033	6750	0.2548	0.9415
0.1805	1.6627	7000	0.2336	0.9358
0.1798	1.7221	7250	0.2671	0.9415
0.1875	1.7815	7500	0.2633	0.9312
0.1957	1.8409	7750	0.2512	0.9404
0.1629	1.9002	8000	0.2643	0.9415
0.1618	1.9596	8250	0.2867	0.9312
0.1463	2.0190	8500	0.3041	0.9369
0.1295	2.0784	8750	0.3021	0.9381
0.1544	2.1378	9000	0.3249	0.9335
0.1704	2.1971	9250	0.3550	0.9346
0.1277	2.2565	9500	0.3159	0.9404
0.157	2.3159	9750	0.3019	0.9381
0.1675	2.3753	10000	0.2930	0.9392
0.1447	2.4347	10250	0.2990	0.9346
0.1658	2.4941	10500	0.2928	0.9346
0.1676	2.5534	10750	0.3317	0.9369
0.1439	2.6128	11000	0.2950	0.9358
0.1827	2.6722	11250	0.2892	0.9369
0.1333	2.7316	11500	0.3014	0.9300
0.1283	2.7910	11750	0.3216	0.9404
0.1349	2.8504	12000	0.3030	0.9369
0.1548	2.9097	12250	0.3232	0.9358
0.1566	2.9691	12500	0.2929	0.9381
0.115	3.0285	12750	0.3122	0.9392
0.1137	3.0879	13000	0.3419	0.9369
0.1471	3.1473	13250	0.3507	0.9392
0.1325	3.2067	13500	0.3332	0.9369
0.1313	3.2660	13750	0.3053	0.9415
0.1471	3.3254	14000	0.2984	0.9346
0.1071	3.3848	14250	0.3294	0.9335
0.1273	3.4442	14500	0.2874	0.9392
0.1203	3.5036	14750	0.3479	0.9392
0.1233	3.5629	15000	0.3529	0.9300
0.0993	3.6223	15250	0.3450	0.9346
0.1525	3.6817	15500	0.3439	0.9335
0.119	3.7411	15750	0.3740	0.9289
0.1081	3.8005	16000	0.3389	0.9381
0.1138	3.8599	16250	0.3383	0.9358
0.123	3.9192	16500	0.3234	0.9369
0.1197	3.9786	16750	0.3422	0.9346
0.1351	4.0380	17000	0.3424	0.9312
0.1006	4.0974	17250	0.3290	0.9392
0.0989	4.1568	17500	0.3456	0.9346
0.1003	4.2162	17750	0.3636	0.9346
0.0911	4.2755	18000	0.3685	0.9346
0.0959	4.3349	18250	0.3547	0.9358
0.1057	4.3943	18500	0.3543	0.9323
0.1135	4.4537	18750	0.3506	0.9346
0.1083	4.5131	19000	0.3361	0.9369
0.0909	4.5724	19250	0.3597	0.9369
0.0966	4.6318	19500	0.3673	0.9369
0.0749	4.6912	19750	0.3665	0.9346
0.0987	4.7506	20000	0.3592	0.9335
0.111	4.8100	20250	0.3531	0.9335
0.0946	4.8694	20500	0.3502	0.9346
0.1023	4.9287	20750	0.3486	0.9346
0.0889	4.9881	21000	0.3491	0.9335

Framework versions

• Transformers 4.52.4
• Pytorch 2.7.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.1