best_model-yelp_polarity-16-100

This model is a fine-tuned version of albert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1862
• Accuracy: 0.8125

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	1	1.0619	0.8438
No log	2.0	2	1.0610	0.8438
No log	3.0	3	1.0591	0.8438
No log	4.0	4	1.0563	0.8438
No log	5.0	5	1.0524	0.8438
No log	6.0	6	1.0473	0.8438
No log	7.0	7	1.0408	0.8438
No log	8.0	8	1.0325	0.8438
No log	9.0	9	1.0221	0.8438
0.5215	10.0	10	1.0093	0.8438
0.5215	11.0	11	0.9939	0.8438
0.5215	12.0	12	0.9775	0.8438
0.5215	13.0	13	0.9630	0.8438
0.5215	14.0	14	0.9517	0.8438
0.5215	15.0	15	0.9431	0.8125
0.5215	16.0	16	0.9352	0.7812
0.5215	17.0	17	0.9263	0.7812
0.5215	18.0	18	0.9195	0.7812
0.5215	19.0	19	0.9178	0.7812
0.3945	20.0	20	0.9230	0.8125
0.3945	21.0	21	0.9374	0.8125
0.3945	22.0	22	0.9628	0.8125
0.3945	23.0	23	1.0035	0.8438
0.3945	24.0	24	1.0608	0.8125
0.3945	25.0	25	1.1258	0.8125
0.3945	26.0	26	1.1859	0.8125
0.3945	27.0	27	1.2311	0.8125
0.3945	28.0	28	1.2580	0.8125
0.3945	29.0	29	1.2702	0.8125
0.2334	30.0	30	1.2750	0.8125
0.2334	31.0	31	1.2763	0.8125
0.2334	32.0	32	1.2763	0.8125
0.2334	33.0	33	1.2757	0.8125
0.2334	34.0	34	1.2733	0.8125
0.2334	35.0	35	1.2687	0.8125
0.2334	36.0	36	1.2612	0.8125
0.2334	37.0	37	1.2508	0.8125
0.2334	38.0	38	1.2376	0.8125
0.2334	39.0	39	1.2213	0.8125
0.024	40.0	40	1.2024	0.8125
0.024	41.0	41	1.1803	0.8125
0.024	42.0	42	1.1548	0.8125
0.024	43.0	43	1.1254	0.8125
0.024	44.0	44	1.0929	0.8125
0.024	45.0	45	1.0591	0.8125
0.024	46.0	46	1.0257	0.8125
0.024	47.0	47	0.9942	0.8125
0.024	48.0	48	0.9662	0.8125
0.024	49.0	49	0.9436	0.8125
0.0008	50.0	50	0.9266	0.8125
0.0008	51.0	51	0.9148	0.8125
0.0008	52.0	52	0.9073	0.8125
0.0008	53.0	53	0.9039	0.8125
0.0008	54.0	54	0.9049	0.8125
0.0008	55.0	55	0.9087	0.8125
0.0008	56.0	56	0.9152	0.8125
0.0008	57.0	57	0.9238	0.8125
0.0008	58.0	58	0.9340	0.8125
0.0008	59.0	59	0.9450	0.8125
0.0006	60.0	60	0.9566	0.8438
0.0006	61.0	61	0.9682	0.8438
0.0006	62.0	62	0.9797	0.8438
0.0006	63.0	63	0.9912	0.8438
0.0006	64.0	64	1.0028	0.8438
0.0006	65.0	65	1.0141	0.8438
0.0006	66.0	66	1.0251	0.8438
0.0006	67.0	67	1.0358	0.8438
0.0006	68.0	68	1.0460	0.8438
0.0006	69.0	69	1.0558	0.8438
0.0005	70.0	70	1.0646	0.8438
0.0005	71.0	71	1.0730	0.8438
0.0005	72.0	72	1.0808	0.8438
0.0005	73.0	73	1.0882	0.8438
0.0005	74.0	74	1.0951	0.8438
0.0005	75.0	75	1.1013	0.8125
0.0005	76.0	76	1.1070	0.8125
0.0005	77.0	77	1.1122	0.8125
0.0005	78.0	78	1.1170	0.8125
0.0005	79.0	79	1.1214	0.8125
0.0004	80.0	80	1.1255	0.8125
0.0004	81.0	81	1.1292	0.8125
0.0004	82.0	82	1.1324	0.8125
0.0004	83.0	83	1.1354	0.8125
0.0004	84.0	84	1.1383	0.8125
0.0004	85.0	85	1.1411	0.8125
0.0004	86.0	86	1.1437	0.8125
0.0004	87.0	87	1.1462	0.8125
0.0004	88.0	88	1.1484	0.8125
0.0004	89.0	89	1.1506	0.8125
0.0004	90.0	90	1.1527	0.8125
0.0004	91.0	91	1.1546	0.8125
0.0004	92.0	92	1.1563	0.8125
0.0004	93.0	93	1.1579	0.8125
0.0004	94.0	94	1.1596	0.8125
0.0004	95.0	95	1.1611	0.8125
0.0004	96.0	96	1.1624	0.8125
0.0004	97.0	97	1.1636	0.8125
0.0004	98.0	98	1.1648	0.8125
0.0004	99.0	99	1.1658	0.8125
0.0003	100.0	100	1.1668	0.8125
0.0003	101.0	101	1.1678	0.8125
0.0003	102.0	102	1.1689	0.8125
0.0003	103.0	103	1.1697	0.8125
0.0003	104.0	104	1.1706	0.8125
0.0003	105.0	105	1.1715	0.8125
0.0003	106.0	106	1.1722	0.8125
0.0003	107.0	107	1.1728	0.8125
0.0003	108.0	108	1.1734	0.8125
0.0003	109.0	109	1.1739	0.8125
0.0003	110.0	110	1.1745	0.8125
0.0003	111.0	111	1.1749	0.8125
0.0003	112.0	112	1.1754	0.8125
0.0003	113.0	113	1.1759	0.8125
0.0003	114.0	114	1.1764	0.8125
0.0003	115.0	115	1.1768	0.8125
0.0003	116.0	116	1.1772	0.8125
0.0003	117.0	117	1.1774	0.8125
0.0003	118.0	118	1.1776	0.8125
0.0003	119.0	119	1.1776	0.8125
0.0003	120.0	120	1.1778	0.8125
0.0003	121.0	121	1.1780	0.8125
0.0003	122.0	122	1.1781	0.8125
0.0003	123.0	123	1.1783	0.8125
0.0003	124.0	124	1.1784	0.8125
0.0003	125.0	125	1.1787	0.8125
0.0003	126.0	126	1.1790	0.8125
0.0003	127.0	127	1.1794	0.8125
0.0003	128.0	128	1.1797	0.8125
0.0003	129.0	129	1.1800	0.8125
0.0003	130.0	130	1.1803	0.8125
0.0003	131.0	131	1.1807	0.8125
0.0003	132.0	132	1.1809	0.8125
0.0003	133.0	133	1.1812	0.8125
0.0003	134.0	134	1.1815	0.8125
0.0003	135.0	135	1.1818	0.8125
0.0003	136.0	136	1.1823	0.8125
0.0003	137.0	137	1.1828	0.8125
0.0003	138.0	138	1.1832	0.8125
0.0003	139.0	139	1.1835	0.8125
0.0002	140.0	140	1.1837	0.8125
0.0002	141.0	141	1.1838	0.8125
0.0002	142.0	142	1.1840	0.8125
0.0002	143.0	143	1.1841	0.8125
0.0002	144.0	144	1.1844	0.8125
0.0002	145.0	145	1.1845	0.8125
0.0002	146.0	146	1.1848	0.8125
0.0002	147.0	147	1.1851	0.8125
0.0002	148.0	148	1.1855	0.8125
0.0002	149.0	149	1.1859	0.8125
0.0002	150.0	150	1.1862	0.8125

Framework versions

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