best_model-sst-2-32-87

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: 1.0406
• Accuracy: 0.8438

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	2	1.2928	0.8438
No log	2.0	4	1.2923	0.8438
No log	3.0	6	1.2917	0.8438
No log	4.0	8	1.2902	0.8438
0.7235	5.0	10	1.2884	0.8438
0.7235	6.0	12	1.2856	0.8438
0.7235	7.0	14	1.2829	0.8438
0.7235	8.0	16	1.2800	0.8281
0.7235	9.0	18	1.2769	0.8281
0.5899	10.0	20	1.2742	0.8281
0.5899	11.0	22	1.2710	0.8281
0.5899	12.0	24	1.2662	0.8281
0.5899	13.0	26	1.2590	0.8281
0.5899	14.0	28	1.2466	0.8281
0.6318	15.0	30	1.2287	0.8281
0.6318	16.0	32	1.2138	0.8281
0.6318	17.0	34	1.2024	0.8281
0.6318	18.0	36	1.1924	0.8281
0.6318	19.0	38	1.1838	0.8281
0.4743	20.0	40	1.1729	0.8281
0.4743	21.0	42	1.1591	0.8281
0.4743	22.0	44	1.1527	0.8281
0.4743	23.0	46	1.1459	0.8281
0.4743	24.0	48	1.1407	0.8281
0.3414	25.0	50	1.1351	0.8281
0.3414	26.0	52	1.1305	0.8281
0.3414	27.0	54	1.1230	0.8281
0.3414	28.0	56	1.1087	0.8281
0.3414	29.0	58	1.0831	0.8281
0.3141	30.0	60	1.0555	0.8281
0.3141	31.0	62	1.0313	0.8438
0.3141	32.0	64	1.0141	0.8594
0.3141	33.0	66	1.0063	0.8438
0.3141	34.0	68	0.9990	0.8438
0.1594	35.0	70	0.9916	0.8438
0.1594	36.0	72	0.9884	0.8438
0.1594	37.0	74	0.9922	0.8438
0.1594	38.0	76	1.0013	0.8281
0.1594	39.0	78	1.0097	0.8281
0.1018	40.0	80	1.0209	0.8281
0.1018	41.0	82	1.0341	0.8281
0.1018	42.0	84	1.0352	0.8281
0.1018	43.0	86	1.0284	0.8281
0.1018	44.0	88	1.0236	0.8281
0.0404	45.0	90	1.0214	0.8438
0.0404	46.0	92	1.0237	0.8594
0.0404	47.0	94	1.0233	0.875
0.0404	48.0	96	1.0223	0.875
0.0404	49.0	98	1.0187	0.875
0.0052	50.0	100	1.0160	0.8594
0.0052	51.0	102	1.0134	0.8594
0.0052	52.0	104	1.0107	0.8438
0.0052	53.0	106	1.0083	0.8438
0.0052	54.0	108	1.0061	0.8438
0.0003	55.0	110	1.0043	0.8438
0.0003	56.0	112	1.0016	0.8438
0.0003	57.0	114	0.9994	0.8438
0.0003	58.0	116	0.9955	0.8438
0.0003	59.0	118	0.9902	0.8438
0.0003	60.0	120	0.9852	0.8438
0.0003	61.0	122	0.9806	0.8438
0.0003	62.0	124	0.9791	0.8438
0.0003	63.0	126	0.9794	0.8438
0.0003	64.0	128	0.9802	0.8438
0.0003	65.0	130	0.9809	0.8438
0.0003	66.0	132	0.9816	0.8438
0.0003	67.0	134	0.9821	0.8438
0.0003	68.0	136	0.9779	0.8438
0.0003	69.0	138	0.9746	0.8281
0.0003	70.0	140	0.9719	0.8281
0.0003	71.0	142	0.9699	0.8281
0.0003	72.0	144	0.9684	0.8438
0.0003	73.0	146	0.9673	0.8438
0.0003	74.0	148	0.9665	0.8438
0.0002	75.0	150	0.9660	0.8438
0.0002	76.0	152	0.9657	0.8438
0.0002	77.0	154	0.9605	0.8438
0.0002	78.0	156	0.9545	0.8438
0.0002	79.0	158	0.9485	0.8438
0.0004	80.0	160	0.9431	0.8438
0.0004	81.0	162	0.9384	0.8438
0.0004	82.0	164	0.9349	0.8438
0.0004	83.0	166	0.9324	0.8438
0.0004	84.0	168	0.9309	0.8438
0.0002	85.0	170	0.9309	0.8438
0.0002	86.0	172	0.9313	0.8438
0.0002	87.0	174	0.9331	0.8438
0.0002	88.0	176	0.9357	0.8438
0.0002	89.0	178	0.9380	0.8438
0.0002	90.0	180	0.9404	0.8438
0.0002	91.0	182	0.9428	0.8438
0.0002	92.0	184	0.9449	0.8438
0.0002	93.0	186	0.9472	0.8438
0.0002	94.0	188	0.9495	0.8438
0.0002	95.0	190	0.9521	0.8438
0.0002	96.0	192	0.9545	0.8438
0.0002	97.0	194	0.9576	0.8438
0.0002	98.0	196	0.9619	0.8438
0.0002	99.0	198	0.9658	0.8438
0.0002	100.0	200	0.9692	0.8438
0.0002	101.0	202	0.9723	0.8438
0.0002	102.0	204	0.9748	0.8438
0.0002	103.0	206	0.9781	0.8438
0.0002	104.0	208	0.9808	0.8438
0.0001	105.0	210	0.9832	0.8438
0.0001	106.0	212	0.9856	0.8438
0.0001	107.0	214	0.9884	0.8438
0.0001	108.0	216	0.9906	0.8438
0.0001	109.0	218	0.9903	0.8438
0.0002	110.0	220	0.9888	0.8438
0.0002	111.0	222	0.9874	0.8438
0.0002	112.0	224	0.9863	0.8438
0.0002	113.0	226	0.9854	0.8438
0.0002	114.0	228	0.9848	0.8438
0.0001	115.0	230	0.9878	0.8438
0.0001	116.0	232	0.9905	0.8438
0.0001	117.0	234	0.9926	0.8438
0.0001	118.0	236	0.9952	0.8438
0.0001	119.0	238	1.0010	0.8438
0.0001	120.0	240	1.0054	0.8438
0.0001	121.0	242	1.0086	0.8438
0.0001	122.0	244	1.0124	0.8438
0.0001	123.0	246	1.0155	0.8438
0.0001	124.0	248	1.0180	0.8438
0.0001	125.0	250	1.0201	0.8438
0.0001	126.0	252	1.0219	0.8438
0.0001	127.0	254	1.0235	0.8438
0.0001	128.0	256	1.0249	0.8438
0.0001	129.0	258	1.0261	0.8438
0.0001	130.0	260	1.0271	0.8438
0.0001	131.0	262	1.0279	0.8438
0.0001	132.0	264	1.0287	0.8438
0.0001	133.0	266	1.0293	0.8438
0.0001	134.0	268	1.0297	0.8438
0.0001	135.0	270	1.0301	0.8438
0.0001	136.0	272	1.0305	0.8438
0.0001	137.0	274	1.0309	0.8438
0.0001	138.0	276	1.0314	0.8438
0.0001	139.0	278	1.0324	0.8438
0.0001	140.0	280	1.0339	0.8438
0.0001	141.0	282	1.0352	0.8438
0.0001	142.0	284	1.0364	0.8438
0.0001	143.0	286	1.0373	0.8438
0.0001	144.0	288	1.0381	0.8438
0.0001	145.0	290	1.0388	0.8438
0.0001	146.0	292	1.0394	0.8438
0.0001	147.0	294	1.0401	0.8438
0.0001	148.0	296	1.0404	0.8438
0.0001	149.0	298	1.0404	0.8438
0.0001	150.0	300	1.0406	0.8438

Framework versions

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