distilbert-base-uncased-finetuned-sarcasm

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

• Loss: 2.3075
• Matthews Correlation: 0.4109

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Matthews Correlation
No log	1.0	57	0.7322	0.0
No log	2.0	114	0.6734	0.1752
No log	3.0	171	0.6436	0.3228
No log	4.0	228	0.7826	0.2778
No log	5.0	285	1.0203	0.2707
No log	6.0	342	1.0190	0.3356
No log	7.0	399	1.1675	0.3177
No log	8.0	456	1.5206	0.2514
0.3597	9.0	513	1.5789	0.4097
0.3597	10.0	570	1.5752	0.3740
0.3597	11.0	627	1.9003	0.3506
0.3597	12.0	684	1.9354	0.3855
0.3597	13.0	741	1.9770	0.3289
0.3597	14.0	798	1.9802	0.3583
0.3597	15.0	855	2.1322	0.3255
0.3597	16.0	912	2.1541	0.2994
0.3597	17.0	969	2.2047	0.2992
0.0329	18.0	1026	2.0794	0.3466
0.0329	19.0	1083	2.0705	0.3012
0.0329	20.0	1140	2.0158	0.3759
0.0329	21.0	1197	2.3999	0.3151
0.0329	22.0	1254	2.1017	0.3917
0.0329	23.0	1311	2.3275	0.3255
0.0329	24.0	1368	2.2258	0.3386
0.0329	25.0	1425	2.3628	0.3406
0.0329	26.0	1482	2.4197	0.3077
0.0145	27.0	1539	2.2661	0.3759
0.0145	28.0	1596	2.4074	0.3077
0.0145	29.0	1653	2.3326	0.3255
0.0145	30.0	1710	2.2813	0.3740
0.0145	31.0	1767	2.3242	0.3181
0.0145	32.0	1824	2.5039	0.2930
0.0145	33.0	1881	2.6045	0.3151
0.0145	34.0	1938	2.3075	0.4109
0.0145	35.0	1995	2.3572	0.3759
0.0129	36.0	2052	2.3833	0.3759
0.0129	37.0	2109	2.6260	0.3009
0.0129	38.0	2166	2.6132	0.3289
0.0129	39.0	2223	2.4151	0.3989
0.0129	40.0	2280	2.5695	0.3360
0.0129	41.0	2337	2.3902	0.3989
0.0129	42.0	2394	2.4388	0.3759
0.0129	43.0	2451	2.6323	0.3289
0.0065	44.0	2508	2.6131	0.3553
0.0065	45.0	2565	2.4426	0.3958
0.0065	46.0	2622	2.4481	0.3958
0.0065	47.0	2679	2.4440	0.3958
0.0065	48.0	2736	2.4689	0.3784
0.0065	49.0	2793	2.4725	0.3784
0.0065	50.0	2850	2.4718	0.3784

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3