scenario-KD-PO-MSV-D2_data-AmazonScience_massive_all_1_155

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: 1.1074
• Accuracy: 0.8669
• F1: 0.8467

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: 55
• 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
1.7059	0.27	5000	1.9979	0.8192	0.7823
1.2894	0.53	10000	1.7859	0.8342	0.8022
1.1214	0.8	15000	1.6643	0.8392	0.8118
0.8728	1.07	20000	1.6370	0.8389	0.8140
0.7934	1.34	25000	1.5969	0.8442	0.8171
0.7466	1.6	30000	1.5752	0.8434	0.8231
0.707	1.87	35000	1.5116	0.8482	0.8282
0.6114	2.14	40000	1.4857	0.8502	0.8282
0.5708	2.41	45000	1.4695	0.8512	0.8297
0.5702	2.67	50000	1.4431	0.8495	0.8277
0.5611	2.94	55000	1.4096	0.8554	0.8356
0.4894	3.21	60000	1.4057	0.8552	0.8353
0.4851	3.47	65000	1.4022	0.8548	0.8325
0.4802	3.74	70000	1.3887	0.8549	0.8339
0.4699	4.01	75000	1.3778	0.8550	0.8359
0.4361	4.28	80000	1.3448	0.8581	0.8379
0.4315	4.54	85000	1.3343	0.8588	0.8341
0.4145	4.81	90000	1.3291	0.8593	0.8361
0.3971	5.08	95000	1.3136	0.8582	0.8371
0.3976	5.34	100000	1.3166	0.8568	0.8398
0.3884	5.61	105000	1.3187	0.8574	0.8361
0.3797	5.88	110000	1.3076	0.8580	0.8352
0.3801	6.15	115000	1.2889	0.8581	0.8381
0.3735	6.41	120000	1.2824	0.8593	0.8394
0.3734	6.68	125000	1.2736	0.8603	0.8399
0.3752	6.95	130000	1.2806	0.8576	0.8341
0.3477	7.22	135000	1.2640	0.8606	0.8405
0.3464	7.48	140000	1.2610	0.8605	0.8390
0.3437	7.75	145000	1.2585	0.8599	0.8379
0.3329	8.02	150000	1.2506	0.8619	0.8399
0.3255	8.28	155000	1.2506	0.8612	0.8398
0.3198	8.55	160000	1.2514	0.8611	0.8400
0.3337	8.82	165000	1.2543	0.8608	0.8416
0.3044	9.09	170000	1.2677	0.8585	0.8406
0.3045	9.35	175000	1.2546	0.8603	0.8394
0.3151	9.62	180000	1.2258	0.8616	0.8420
0.3091	9.89	185000	1.2356	0.8618	0.8409
0.2889	10.15	190000	1.2244	0.8613	0.8417
0.2931	10.42	195000	1.2106	0.8623	0.8423
0.2923	10.69	200000	1.2272	0.8611	0.8409
0.2988	10.96	205000	1.2070	0.8632	0.8418
0.2817	11.22	210000	1.2079	0.8624	0.8429
0.2878	11.49	215000	1.2132	0.8633	0.8410
0.2803	11.76	220000	1.2023	0.8619	0.8428
0.2769	12.03	225000	1.2024	0.8621	0.8438
0.2807	12.29	230000	1.1938	0.8632	0.8434
0.2795	12.56	235000	1.2024	0.8632	0.8417
0.277	12.83	240000	1.1924	0.8623	0.8419
0.2602	13.09	245000	1.1960	0.8623	0.8424
0.268	13.36	250000	1.1893	0.8617	0.8427
0.2653	13.63	255000	1.1890	0.8620	0.8394
0.2558	13.9	260000	1.1790	0.8634	0.8422
0.2602	14.16	265000	1.1760	0.8645	0.8429
0.256	14.43	270000	1.1714	0.8635	0.8442
0.2463	14.7	275000	1.1855	0.8626	0.8421
0.2546	14.96	280000	1.1791	0.8640	0.8439
0.2499	15.23	285000	1.1763	0.8640	0.8451
0.2539	15.5	290000	1.1693	0.8643	0.8447
0.2466	15.77	295000	1.1607	0.8646	0.8444
0.2376	16.03	300000	1.1665	0.8637	0.8427
0.2397	16.3	305000	1.1754	0.8639	0.8441
0.2408	16.57	310000	1.1732	0.8639	0.8437
0.2443	16.84	315000	1.1621	0.8631	0.8421
0.2273	17.1	320000	1.1572	0.8646	0.8447
0.2314	17.37	325000	1.1578	0.8643	0.8438
0.2376	17.64	330000	1.1571	0.8644	0.8434
0.2296	17.9	335000	1.1504	0.8657	0.8470
0.2254	18.17	340000	1.1542	0.8640	0.8435
0.2305	18.44	345000	1.1599	0.8640	0.8427
0.2236	18.71	350000	1.1566	0.8638	0.8439
0.2276	18.97	355000	1.1425	0.8661	0.8469
0.2223	19.24	360000	1.1580	0.8648	0.8454
0.2242	19.51	365000	1.1406	0.8651	0.8455
0.2235	19.77	370000	1.1490	0.8652	0.8455
0.2183	20.04	375000	1.1342	0.8652	0.8451
0.2123	20.31	380000	1.1457	0.8649	0.8443
0.2162	20.58	385000	1.1328	0.8655	0.8452
0.2111	20.84	390000	1.1362	0.8657	0.8450
0.2121	21.11	395000	1.1349	0.8655	0.8450
0.204	21.38	400000	1.1332	0.8651	0.8447
0.2133	21.65	405000	1.1330	0.8642	0.8438
0.2115	21.91	410000	1.1339	0.8647	0.8440
0.2054	22.18	415000	1.1316	0.8647	0.8444
0.211	22.45	420000	1.1286	0.8660	0.8452
0.2015	22.71	425000	1.1290	0.8656	0.8462
0.2112	22.98	430000	1.1342	0.8654	0.8450
0.2016	23.25	435000	1.1288	0.8650	0.8453
0.1991	23.52	440000	1.1303	0.8657	0.8468
0.1988	23.78	445000	1.1238	0.8658	0.8463
0.1955	24.05	450000	1.1189	0.8664	0.8471
0.1964	24.32	455000	1.1254	0.8655	0.8447
0.2009	24.58	460000	1.1209	0.8659	0.8468
0.1999	24.85	465000	1.1166	0.8657	0.8449
0.1871	25.12	470000	1.1251	0.8657	0.8449
0.1934	25.39	475000	1.1145	0.8657	0.8446
0.1933	25.65	480000	1.1187	0.8651	0.8453
0.197	25.92	485000	1.1188	0.8658	0.8448
0.1938	26.19	490000	1.1176	0.8657	0.8455
0.196	26.46	495000	1.1221	0.8660	0.8459
0.1868	26.72	500000	1.1166	0.8660	0.8454
0.1917	26.99	505000	1.1148	0.8668	0.8468
0.1893	27.26	510000	1.1130	0.8660	0.8457
0.1881	27.52	515000	1.1138	0.8658	0.8454
0.1857	27.79	520000	1.1139	0.8662	0.8457
0.1904	28.06	525000	1.1112	0.8657	0.8457
0.186	28.33	530000	1.1121	0.8664	0.8454
0.186	28.59	535000	1.1115	0.8665	0.8471
0.1866	28.86	540000	1.1091	0.8659	0.8450
0.1847	29.13	545000	1.1094	0.8668	0.8467
0.1826	29.39	550000	1.1116	0.8662	0.8463
0.1831	29.66	555000	1.1085	0.8660	0.8459
0.1819	29.93	560000	1.1074	0.8669	0.8467

Framework versions

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