scenario-KD-PO-MSV-D2_data-AmazonScience_massive_all_1_144

This model is a fine-tuned version of haryoaw/scenario-MDBT-TCR_data-AmazonScience_massive_all_1_1 on the massive dataset. It achieves the following results on the evaluation set:

• Loss: 1.1139
• Accuracy: 0.8659
• F1: 0.8466

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 44
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.7627	0.27	5000	1.9858	0.8200	0.7785
1.3268	0.53	10000	1.7497	0.8353	0.8042
1.1431	0.8	15000	1.6427	0.8403	0.8129
0.846	1.07	20000	1.5717	0.8445	0.8134
0.7952	1.34	25000	1.5650	0.8457	0.8207
0.7416	1.6	30000	1.5495	0.8493	0.8246
0.713	1.87	35000	1.4702	0.8518	0.8325
0.6101	2.14	40000	1.4505	0.8522	0.8287
0.591	2.41	45000	1.4639	0.8522	0.8257
0.5774	2.67	50000	1.4352	0.8498	0.8292
0.5564	2.94	55000	1.4054	0.8556	0.8357
0.48	3.21	60000	1.4085	0.8539	0.8330
0.4767	3.47	65000	1.3640	0.8578	0.8352
0.4725	3.74	70000	1.3570	0.8553	0.8326
0.4599	4.01	75000	1.3478	0.8568	0.8356
0.4328	4.28	80000	1.3349	0.8561	0.8329
0.4237	4.54	85000	1.3403	0.8585	0.8375
0.4186	4.81	90000	1.3329	0.8591	0.8402
0.3964	5.08	95000	1.3430	0.8552	0.8350
0.3896	5.34	100000	1.3224	0.8572	0.8369
0.4012	5.61	105000	1.3126	0.8597	0.8389
0.3845	5.88	110000	1.3096	0.8574	0.8368
0.3601	6.15	115000	1.2671	0.8597	0.8378
0.3608	6.41	120000	1.2839	0.8596	0.8374
0.3612	6.68	125000	1.2874	0.8590	0.8393
0.3617	6.95	130000	1.3028	0.8572	0.8382
0.3357	7.22	135000	1.2707	0.8597	0.8397
0.3467	7.48	140000	1.2735	0.8597	0.8407
0.3516	7.75	145000	1.2572	0.8606	0.8397
0.3298	8.02	150000	1.2609	0.8601	0.8396
0.3258	8.28	155000	1.2527	0.8585	0.8381
0.3301	8.55	160000	1.2414	0.8613	0.8445
0.3193	8.82	165000	1.2525	0.8602	0.8407
0.307	9.09	170000	1.2398	0.8613	0.8428
0.3155	9.35	175000	1.2209	0.8621	0.8416
0.3138	9.62	180000	1.2155	0.8617	0.8428
0.3108	9.89	185000	1.2363	0.8618	0.8420
0.2954	10.15	190000	1.2141	0.8609	0.8414
0.2968	10.42	195000	1.2331	0.8611	0.8430
0.3017	10.69	200000	1.2081	0.8618	0.8433
0.2989	10.96	205000	1.2025	0.8630	0.8438
0.2792	11.22	210000	1.2063	0.8635	0.8433
0.2946	11.49	215000	1.1898	0.8639	0.8435
0.2778	11.76	220000	1.2013	0.8625	0.8428
0.2757	12.03	225000	1.1908	0.8627	0.8439
0.2763	12.29	230000	1.1906	0.8631	0.8424
0.2699	12.56	235000	1.1894	0.8629	0.8422
0.2716	12.83	240000	1.1887	0.8643	0.8460
0.2715	13.09	245000	1.1940	0.8634	0.8463
0.2659	13.36	250000	1.1844	0.8642	0.8443
0.2693	13.63	255000	1.1844	0.8642	0.8446
0.2654	13.9	260000	1.1784	0.8637	0.8454
0.2522	14.16	265000	1.1728	0.8641	0.8433
0.2604	14.43	270000	1.1878	0.8627	0.8415
0.2489	14.7	275000	1.1795	0.8640	0.8437
0.2611	14.96	280000	1.1585	0.8648	0.8450
0.246	15.23	285000	1.1574	0.8647	0.8452
0.2482	15.5	290000	1.1654	0.8633	0.8427
0.2423	15.77	295000	1.1683	0.8632	0.8416
0.2387	16.03	300000	1.1736	0.8625	0.8415
0.2417	16.3	305000	1.1686	0.8635	0.8437
0.2433	16.57	310000	1.1639	0.8626	0.8410
0.2405	16.84	315000	1.1647	0.8632	0.8421
0.2327	17.1	320000	1.1459	0.8642	0.8446
0.2374	17.37	325000	1.1513	0.8643	0.8454
0.233	17.64	330000	1.1479	0.8649	0.8438
0.2381	17.9	335000	1.1556	0.8643	0.8450
0.2228	18.17	340000	1.1523	0.8648	0.8461
0.2305	18.44	345000	1.1523	0.8635	0.8440
0.2244	18.71	350000	1.1507	0.8648	0.8450
0.2212	18.97	355000	1.1413	0.8650	0.8451
0.2207	19.24	360000	1.1401	0.8644	0.8466
0.2164	19.51	365000	1.1518	0.8630	0.8426
0.2232	19.77	370000	1.1469	0.8640	0.8452
0.2147	20.04	375000	1.1495	0.8629	0.8433
0.2089	20.31	380000	1.1392	0.8657	0.8457
0.2074	20.58	385000	1.1381	0.8643	0.8441
0.2149	20.84	390000	1.1415	0.8639	0.8447
0.2074	21.11	395000	1.1307	0.8647	0.8441
0.2087	21.38	400000	1.1351	0.8641	0.8432
0.2104	21.65	405000	1.1312	0.8644	0.8448
0.2078	21.91	410000	1.1296	0.8650	0.8457
0.2038	22.18	415000	1.1249	0.8657	0.8452
0.2037	22.45	420000	1.1334	0.8643	0.8444
0.2027	22.71	425000	1.1280	0.8644	0.8446
0.2041	22.98	430000	1.1321	0.8640	0.8467
0.1921	23.25	435000	1.1324	0.8635	0.8445
0.2007	23.52	440000	1.1248	0.8649	0.8458
0.2018	23.78	445000	1.1274	0.8649	0.8451
0.2009	24.05	450000	1.1204	0.8654	0.8473
0.2017	24.32	455000	1.1245	0.8651	0.8450
0.1977	24.58	460000	1.1234	0.8649	0.8461
0.1974	24.85	465000	1.1234	0.8646	0.8445
0.1934	25.12	470000	1.1220	0.8655	0.8457
0.1937	25.39	475000	1.1183	0.8659	0.8470
0.1947	25.65	480000	1.1207	0.8653	0.8449
0.1932	25.92	485000	1.1172	0.8660	0.8465
0.1904	26.19	490000	1.1250	0.8651	0.8462
0.1892	26.46	495000	1.1171	0.8661	0.8470
0.196	26.72	500000	1.1207	0.8652	0.8456
0.1927	26.99	505000	1.1174	0.8651	0.8454
0.1888	27.26	510000	1.1183	0.8653	0.8457
0.1906	27.52	515000	1.1190	0.8646	0.8455
0.1854	27.79	520000	1.1098	0.8655	0.8450
0.1869	28.06	525000	1.1173	0.8658	0.8463
0.1829	28.33	530000	1.1157	0.8654	0.8453
0.1909	28.59	535000	1.1135	0.8655	0.8459
0.1853	28.86	540000	1.1103	0.8663	0.8475
0.1847	29.13	545000	1.1141	0.8651	0.8455
0.1814	29.39	550000	1.1101	0.8655	0.8456
0.1794	29.66	555000	1.1113	0.8657	0.8460
0.1842	29.93	560000	1.1139	0.8659	0.8466

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3