scenario-KD-PO-MSV-D2_data-AmazonScience_massive_all_1_166

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: 1.1092
• Accuracy: 0.8656
• F1: 0.8457

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: 66
• 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
1.6991	0.27	5000	2.0126	0.8189	0.7729
1.3038	0.53	10000	1.7765	0.8309	0.7954
1.1117	0.8	15000	1.6835	0.8384	0.8084
0.7956	1.07	20000	1.6250	0.8465	0.8119
0.7732	1.34	25000	1.6081	0.8452	0.8183
0.7648	1.6	30000	1.5203	0.8492	0.8311
0.729	1.87	35000	1.4829	0.8494	0.8247
0.5929	2.14	40000	1.4956	0.8493	0.8304
0.5728	2.41	45000	1.4492	0.8530	0.8312
0.5387	2.67	50000	1.4214	0.8527	0.8325
0.5635	2.94	55000	1.4261	0.8508	0.8290
0.5014	3.21	60000	1.4044	0.8543	0.8337
0.4891	3.47	65000	1.3848	0.8552	0.8330
0.476	3.74	70000	1.3737	0.8548	0.8338
0.4534	4.01	75000	1.3473	0.8571	0.8357
0.431	4.28	80000	1.3566	0.8571	0.8353
0.4351	4.54	85000	1.3364	0.8584	0.8372
0.4385	4.81	90000	1.3211	0.8576	0.8350
0.3919	5.08	95000	1.3317	0.8552	0.8333
0.3944	5.34	100000	1.3159	0.8590	0.8392
0.3961	5.61	105000	1.3136	0.8579	0.8382
0.3923	5.88	110000	1.3012	0.8589	0.8387
0.3597	6.15	115000	1.2809	0.8585	0.8346
0.3675	6.41	120000	1.2807	0.8592	0.8397
0.3692	6.68	125000	1.2721	0.8591	0.8383
0.3661	6.95	130000	1.2968	0.8591	0.8373
0.3473	7.22	135000	1.2938	0.8572	0.8360
0.3412	7.48	140000	1.2709	0.8586	0.8390
0.3375	7.75	145000	1.2589	0.8597	0.8389
0.3317	8.02	150000	1.2484	0.8609	0.8425
0.3296	8.28	155000	1.2351	0.8621	0.8416
0.32	8.55	160000	1.2354	0.8621	0.8400
0.3231	8.82	165000	1.2361	0.8594	0.8394
0.3137	9.09	170000	1.2152	0.8642	0.8449
0.301	9.35	175000	1.2443	0.8594	0.8378
0.3106	9.62	180000	1.2355	0.8609	0.8414
0.308	9.89	185000	1.2301	0.8615	0.8437
0.3045	10.15	190000	1.2243	0.8611	0.8399
0.2915	10.42	195000	1.2303	0.8593	0.8404
0.2971	10.69	200000	1.2041	0.8617	0.8406
0.2925	10.96	205000	1.2160	0.8616	0.8415
0.2822	11.22	210000	1.2128	0.8602	0.8407
0.2868	11.49	215000	1.2180	0.8600	0.8404
0.2844	11.76	220000	1.1990	0.8639	0.8438
0.2718	12.03	225000	1.2098	0.8597	0.8385
0.2734	12.29	230000	1.2064	0.8601	0.8396
0.2718	12.56	235000	1.2063	0.8621	0.8426
0.2666	12.83	240000	1.1999	0.8614	0.8411
0.2656	13.09	245000	1.1774	0.8629	0.8429
0.2614	13.36	250000	1.1894	0.8622	0.8430
0.2702	13.63	255000	1.1898	0.8635	0.8443
0.2631	13.9	260000	1.1805	0.8636	0.8446
0.2464	14.16	265000	1.1690	0.8634	0.8425
0.2505	14.43	270000	1.1708	0.8640	0.8438
0.2575	14.7	275000	1.1691	0.8628	0.8416
0.2592	14.96	280000	1.1865	0.8620	0.8418
0.2476	15.23	285000	1.1667	0.8634	0.8402
0.244	15.5	290000	1.1691	0.8635	0.8444
0.2398	15.77	295000	1.1654	0.8632	0.8444
0.2406	16.03	300000	1.1637	0.8650	0.8460
0.2319	16.3	305000	1.1713	0.8635	0.8430
0.2419	16.57	310000	1.1671	0.8636	0.8463
0.2352	16.84	315000	1.1590	0.8642	0.8446
0.2337	17.1	320000	1.1515	0.8645	0.8439
0.2364	17.37	325000	1.1580	0.8638	0.8431
0.2302	17.64	330000	1.1606	0.8636	0.8424
0.2308	17.9	335000	1.1573	0.8636	0.8441
0.2306	18.17	340000	1.1477	0.8643	0.8434
0.2208	18.44	345000	1.1556	0.8646	0.8435
0.2273	18.71	350000	1.1611	0.8632	0.8427
0.2272	18.97	355000	1.1514	0.8637	0.8454
0.219	19.24	360000	1.1405	0.8650	0.8458
0.2186	19.51	365000	1.1509	0.8645	0.8448
0.2268	19.77	370000	1.1432	0.8659	0.8472
0.2129	20.04	375000	1.1417	0.8648	0.8462
0.2192	20.31	380000	1.1360	0.8643	0.8451
0.2114	20.58	385000	1.1454	0.8640	0.8440
0.216	20.84	390000	1.1384	0.8639	0.8447
0.2157	21.11	395000	1.1516	0.8636	0.8463
0.2065	21.38	400000	1.1342	0.8646	0.8447
0.2089	21.65	405000	1.1315	0.8646	0.8447
0.2136	21.91	410000	1.1371	0.8644	0.8454
0.2024	22.18	415000	1.1354	0.8642	0.8458
0.2025	22.45	420000	1.1343	0.8639	0.8440
0.2015	22.71	425000	1.1353	0.8646	0.8455
0.2049	22.98	430000	1.1287	0.8651	0.8450
0.1957	23.25	435000	1.1265	0.8651	0.8451
0.2049	23.52	440000	1.1294	0.8644	0.8441
0.2019	23.78	445000	1.1233	0.8660	0.8479
0.1982	24.05	450000	1.1237	0.8653	0.8466
0.1946	24.32	455000	1.1233	0.8655	0.8474
0.1982	24.58	460000	1.1264	0.8660	0.8484
0.1963	24.85	465000	1.1269	0.8647	0.8466
0.1976	25.12	470000	1.1245	0.8649	0.8463
0.1955	25.39	475000	1.1145	0.8655	0.8454
0.1931	25.65	480000	1.1145	0.8655	0.8466
0.193	25.92	485000	1.1114	0.8659	0.8465
0.194	26.19	490000	1.1109	0.8655	0.8455
0.1935	26.46	495000	1.1172	0.8652	0.8451
0.1869	26.72	500000	1.1175	0.8657	0.8460
0.1907	26.99	505000	1.1174	0.8649	0.8446
0.1899	27.26	510000	1.1173	0.8653	0.8454
0.1865	27.52	515000	1.1164	0.8656	0.8459
0.1904	27.79	520000	1.1124	0.8659	0.8457
0.1888	28.06	525000	1.1136	0.8649	0.8452
0.1812	28.33	530000	1.1166	0.8653	0.8459
0.1871	28.59	535000	1.1133	0.8656	0.8453
0.1857	28.86	540000	1.1149	0.8651	0.8457
0.1844	29.13	545000	1.1139	0.8657	0.8461
0.1864	29.39	550000	1.1158	0.8649	0.8453
0.1779	29.66	555000	1.1102	0.8656	0.8462
0.1816	29.93	560000	1.1092	0.8656	0.8457

Framework versions

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