scenario-KD-SCR-MSV-EN-EN-D2_data-en-massive_all_1_166

This model is a fine-tuned version of haryoaw/scenario-MDBT-TCR_data-en-massive_all_1_1 on the massive dataset. It achieves the following results on the evaluation set:

• Loss: 303.6620
• Accuracy: 0.0845
• F1: 0.0403

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: 8
• eval_batch_size: 32
• seed: 66
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• 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
No log	0.2778	100	628.3530	0.0645	0.0031
No log	0.5556	200	607.2537	0.0645	0.0021
No log	0.8333	300	597.3507	0.0644	0.0021
No log	1.1111	400	589.9023	0.0646	0.0023
540.0628	1.3889	500	587.1896	0.0669	0.0032
540.0628	1.6667	600	574.1956	0.0650	0.0024
540.0628	1.9444	700	567.6414	0.0649	0.0024
540.0628	2.2222	800	561.3796	0.0673	0.0033
540.0628	2.5	900	553.5518	0.0693	0.0037
428.1685	2.7778	1000	547.9289	0.0755	0.0058
428.1685	3.0556	1100	543.2942	0.0754	0.0063
428.1685	3.3333	1200	537.8929	0.0738	0.0050
428.1685	3.6111	1300	530.7917	0.0539	0.0050
428.1685	3.8889	1400	523.8170	0.0725	0.0066
372.0595	4.1667	1500	515.7302	0.0569	0.0045
372.0595	4.4444	1600	512.6782	0.0687	0.0076
372.0595	4.7222	1700	504.7878	0.0691	0.0056
372.0595	5.0	1800	498.9977	0.0593	0.0068
372.0595	5.2778	1900	495.0116	0.0692	0.0093
329.4533	5.5556	2000	490.4921	0.0549	0.0073
329.4533	5.8333	2100	485.8165	0.0782	0.0108
329.4533	6.1111	2200	481.0579	0.0677	0.0110
329.4533	6.3889	2300	475.8662	0.0652	0.0119
329.4533	6.6667	2400	469.8492	0.0618	0.0107
297.8759	6.9444	2500	468.4412	0.0727	0.0132
297.8759	7.2222	2600	461.2780	0.0596	0.0120
297.8759	7.5	2700	456.7401	0.0662	0.0140
297.8759	7.7778	2800	456.5062	0.0587	0.0124
297.8759	8.0556	2900	449.0881	0.0583	0.0116
271.9938	8.3333	3000	443.4659	0.0651	0.0155
271.9938	8.6111	3100	439.0943	0.0768	0.0175
271.9938	8.8889	3200	436.8179	0.0738	0.0153
271.9938	9.1667	3300	431.4831	0.0666	0.0179
271.9938	9.4444	3400	427.3849	0.0659	0.0171
249.2916	9.7222	3500	421.9467	0.0588	0.0134
249.2916	10.0	3600	419.9226	0.0657	0.0166
249.2916	10.2778	3700	414.5518	0.0724	0.0194
249.2916	10.5556	3800	414.6526	0.0693	0.0193
249.2916	10.8333	3900	410.1772	0.0746	0.0218
229.8694	11.1111	4000	407.4032	0.0791	0.0250
229.8694	11.3889	4100	401.9918	0.0747	0.0248
229.8694	11.6667	4200	399.7811	0.0596	0.0162
229.8694	11.9444	4300	396.3000	0.0675	0.0222
229.8694	12.2222	4400	394.0766	0.0717	0.0241
212.7051	12.5	4500	389.2682	0.0793	0.0294
212.7051	12.7778	4600	387.6005	0.0703	0.0237
212.7051	13.0556	4700	385.7607	0.0785	0.0275
212.7051	13.3333	4800	380.1237	0.0755	0.0292
212.7051	13.6111	4900	380.8004	0.0722	0.0262
198.1273	13.8889	5000	373.5923	0.0800	0.0317
198.1273	14.1667	5100	373.1263	0.0719	0.0266
198.1273	14.4444	5200	368.4973	0.0698	0.0282
198.1273	14.7222	5300	368.4087	0.0772	0.0295
198.1273	15.0	5400	365.3843	0.0757	0.0300
184.8348	15.2778	5500	360.9421	0.0743	0.0309
184.8348	15.5556	5600	359.4719	0.0806	0.0339
184.8348	15.8333	5700	356.0381	0.0692	0.0283
184.8348	16.1111	5800	356.5358	0.0831	0.0342
184.8348	16.3889	5900	352.8613	0.0811	0.0353
173.7237	16.6667	6000	350.3405	0.0788	0.0340
173.7237	16.9444	6100	349.9302	0.0763	0.0306
173.7237	17.2222	6200	345.5006	0.0842	0.0378
173.7237	17.5	6300	343.7187	0.0778	0.0343
173.7237	17.7778	6400	341.6388	0.0785	0.0353
164.1808	18.0556	6500	341.8687	0.0866	0.0380
164.1808	18.3333	6600	340.1183	0.0843	0.0375
164.1808	18.6111	6700	336.7932	0.0833	0.0373
164.1808	18.8889	6800	334.5229	0.0811	0.0333
164.1808	19.1667	6900	334.8410	0.0896	0.0395
155.9113	19.4444	7000	335.9225	0.0834	0.0352
155.9113	19.7222	7100	332.2399	0.0832	0.0351
155.9113	20.0	7200	328.4766	0.0803	0.0355
155.9113	20.2778	7300	327.9207	0.0840	0.0400
155.9113	20.5556	7400	325.6150	0.0813	0.0379
148.9894	20.8333	7500	324.3223	0.0796	0.0358
148.9894	21.1111	7600	323.8501	0.0803	0.0368
148.9894	21.3889	7700	322.7423	0.0847	0.0392
148.9894	21.6667	7800	321.7413	0.0833	0.0383
148.9894	21.9444	7900	320.6828	0.0767	0.0358
142.9619	22.2222	8000	319.8029	0.0842	0.0370
142.9619	22.5	8100	317.4704	0.0889	0.0417
142.9619	22.7778	8200	316.1092	0.0840	0.0392
142.9619	23.0556	8300	316.0903	0.0871	0.0410
142.9619	23.3333	8400	314.5365	0.0826	0.0383
138.3276	23.6111	8500	313.4515	0.0823	0.0390
138.3276	23.8889	8600	313.9167	0.0850	0.0398
138.3276	24.1667	8700	312.8665	0.0815	0.0384
138.3276	24.4444	8800	310.8644	0.0819	0.0387
138.3276	24.7222	8900	310.7165	0.0846	0.0396
134.11	25.0	9000	311.4012	0.0880	0.0399
134.11	25.2778	9100	308.5638	0.0852	0.0401
134.11	25.5556	9200	308.5912	0.0843	0.0405
134.11	25.8333	9300	308.6916	0.0855	0.0399
134.11	26.1111	9400	307.4913	0.0872	0.0421
131.1968	26.3889	9500	306.1313	0.0825	0.0393
131.1968	26.6667	9600	306.2299	0.0806	0.0384
131.1968	26.9444	9700	306.1326	0.0842	0.0398
131.1968	27.2222	9800	305.8682	0.0856	0.0398
131.1968	27.5	9900	305.3973	0.0826	0.0398
128.8904	27.7778	10000	305.0905	0.0816	0.0392
128.8904	28.0556	10100	304.0052	0.0828	0.0394
128.8904	28.3333	10200	303.5638	0.0855	0.0413
128.8904	28.6111	10300	304.2411	0.0830	0.0395
128.8904	28.8889	10400	303.9249	0.0842	0.0398
127.4357	29.1667	10500	304.0102	0.0833	0.0392
127.4357	29.4444	10600	304.0796	0.0842	0.0406
127.4357	29.7222	10700	303.9043	0.0846	0.0403
127.4357	30.0	10800	303.6620	0.0845	0.0403

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1