best_model-sst-2-64-21

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

• Loss: 1.0374
• Accuracy: 0.8906

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 150

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	4	1.1068	0.8672
No log	2.0	8	1.1055	0.8672
0.5789	3.0	12	1.1002	0.8672
0.5789	4.0	16	1.0902	0.8672
0.4952	5.0	20	1.0797	0.8672
0.4952	6.0	24	1.0662	0.8672
0.4952	7.0	28	1.0461	0.8672
0.4202	8.0	32	1.0329	0.8672
0.4202	9.0	36	1.0326	0.8672
0.5159	10.0	40	1.0217	0.8672
0.5159	11.0	44	1.0053	0.8672
0.5159	12.0	48	0.9908	0.875
0.4018	13.0	52	0.9818	0.8828
0.4018	14.0	56	0.9686	0.8828
0.2452	15.0	60	0.9591	0.8828
0.2452	16.0	64	0.9489	0.8828
0.2452	17.0	68	0.9421	0.8828
0.1966	18.0	72	0.9354	0.8828
0.1966	19.0	76	0.9318	0.8906
0.1955	20.0	80	0.9353	0.8828
0.1955	21.0	84	0.9552	0.8828
0.1955	22.0	88	0.9728	0.875
0.1316	23.0	92	0.9686	0.875
0.1316	24.0	96	0.9555	0.875
0.0488	25.0	100	0.9442	0.8828
0.0488	26.0	104	0.9410	0.8828
0.0488	27.0	108	0.9413	0.8828
0.0023	28.0	112	0.9522	0.8828
0.0023	29.0	116	0.9614	0.8828
0.0019	30.0	120	0.9603	0.8828
0.0019	31.0	124	0.9474	0.8828
0.0019	32.0	128	0.9408	0.8906
0.0136	33.0	132	0.9417	0.8906
0.0136	34.0	136	0.9433	0.8906
0.0037	35.0	140	0.9412	0.8906
0.0037	36.0	144	0.9529	0.8906
0.0037	37.0	148	0.9641	0.8828
0.0003	38.0	152	0.9868	0.8828
0.0003	39.0	156	0.9985	0.875
0.0002	40.0	160	1.0006	0.875
0.0002	41.0	164	1.0009	0.875
0.0002	42.0	168	1.0038	0.875
0.0013	43.0	172	0.9982	0.8828
0.0013	44.0	176	0.9853	0.8828
0.0102	45.0	180	0.9790	0.8828
0.0102	46.0	184	0.9900	0.8828
0.0102	47.0	188	1.0004	0.8828
0.0002	48.0	192	1.0063	0.875
0.0002	49.0	196	1.0095	0.875
0.0001	50.0	200	1.0136	0.875
0.0001	51.0	204	1.0180	0.8672
0.0001	52.0	208	1.0206	0.8672
0.0001	53.0	212	1.0178	0.8672
0.0001	54.0	216	1.0157	0.8672
0.0001	55.0	220	1.0140	0.875
0.0001	56.0	224	1.0128	0.875
0.0001	57.0	228	1.0117	0.875
0.0001	58.0	232	1.0097	0.875
0.0001	59.0	236	1.0082	0.875
0.0001	60.0	240	1.0002	0.8828
0.0001	61.0	244	0.9944	0.8828
0.0001	62.0	248	0.9913	0.8906
0.0001	63.0	252	0.9897	0.8906
0.0001	64.0	256	0.9893	0.8906
0.0001	65.0	260	0.9895	0.8906
0.0001	66.0	264	0.9899	0.8906
0.0001	67.0	268	0.9905	0.8906
0.0001	68.0	272	0.9913	0.8906
0.0001	69.0	276	0.9962	0.8906
0.0001	70.0	280	1.0023	0.8828
0.0001	71.0	284	1.0079	0.8828
0.0001	72.0	288	1.0118	0.875
0.0001	73.0	292	1.0144	0.875
0.0001	74.0	296	1.0161	0.875
0.0001	75.0	300	1.0172	0.875
0.0001	76.0	304	1.0178	0.875
0.0001	77.0	308	1.0241	0.875
0.0183	78.0	312	1.0549	0.8672
0.0183	79.0	316	1.0631	0.8672
0.0001	80.0	320	1.0629	0.875
0.0001	81.0	324	1.0650	0.875
0.0001	82.0	328	1.0672	0.8594
0.0001	83.0	332	1.0686	0.8594
0.0001	84.0	336	1.0632	0.875
0.0131	85.0	340	1.0157	0.8672
0.0131	86.0	344	0.9959	0.8828
0.0131	87.0	348	0.9946	0.8906
0.0001	88.0	352	0.9933	0.8906
0.0001	89.0	356	0.9933	0.8906
0.0001	90.0	360	0.9941	0.8828
0.0001	91.0	364	0.9949	0.8828
0.0001	92.0	368	0.9954	0.8828
0.0001	93.0	372	0.9959	0.8828
0.0001	94.0	376	0.9962	0.8828
0.0001	95.0	380	0.9961	0.8828
0.0001	96.0	384	0.9963	0.8828
0.0001	97.0	388	0.9967	0.8828
0.0001	98.0	392	0.9987	0.8906
0.0001	99.0	396	1.0214	0.8828
0.0105	100.0	400	1.0346	0.875
0.0105	101.0	404	1.0406	0.875
0.0105	102.0	408	1.0435	0.875
0.0001	103.0	412	1.0444	0.875
0.0001	104.0	416	1.0446	0.875
0.0001	105.0	420	1.0447	0.875
0.0001	106.0	424	1.0448	0.875
0.0001	107.0	428	1.0453	0.8828
0.0001	108.0	432	1.0457	0.8828
0.0001	109.0	436	1.0488	0.875
0.0184	110.0	440	1.0597	0.875
0.0184	111.0	444	1.0939	0.8594
0.0184	112.0	448	1.1410	0.8438
0.0001	113.0	452	1.1659	0.8438
0.0001	114.0	456	1.1104	0.8594
0.0001	115.0	460	1.0816	0.8672
0.0001	116.0	464	1.0695	0.875
0.0001	117.0	468	1.0702	0.875
0.0	118.0	472	1.0709	0.875
0.0	119.0	476	1.0704	0.875
0.0	120.0	480	1.0693	0.875
0.0	121.0	484	1.0684	0.875
0.0	122.0	488	1.0677	0.875
0.0	123.0	492	1.0676	0.875
0.0	124.0	496	1.0676	0.875
0.0	125.0	500	1.0675	0.875
0.0	126.0	504	1.0675	0.875
0.0	127.0	508	1.0676	0.875
0.0	128.0	512	1.0687	0.875
0.0	129.0	516	1.0694	0.875
0.0	130.0	520	1.0701	0.875
0.0	131.0	524	1.0707	0.875
0.0	132.0	528	1.0661	0.875
0.0001	133.0	532	1.0391	0.8906
0.0001	134.0	536	1.0258	0.8906
0.0	135.0	540	1.0188	0.8906
0.0	136.0	544	1.0171	0.8906
0.0	137.0	548	1.0188	0.8828
0.0	138.0	552	1.0210	0.875
0.0	139.0	556	1.0223	0.875
0.0001	140.0	560	1.0202	0.8828
0.0001	141.0	564	1.0235	0.8906
0.0001	142.0	568	1.0288	0.8906
0.0	143.0	572	1.0322	0.8906
0.0	144.0	576	1.0343	0.8906
0.0	145.0	580	1.0356	0.8906
0.0	146.0	584	1.0364	0.8906
0.0	147.0	588	1.0369	0.8906
0.0	148.0	592	1.0372	0.8906
0.0	149.0	596	1.0374	0.8906
0.0	150.0	600	1.0374	0.8906

Framework versions

• Transformers 4.32.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.4.0
• Tokenizers 0.13.3