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

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

• Loss: 306.4669
• Accuracy: 0.0905
• F1: 0.0439

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	627.7275	0.0646	0.0058
No log	0.5556	200	607.2649	0.0644	0.0021
No log	0.8333	300	599.7544	0.0652	0.0026
No log	1.1111	400	589.6091	0.0679	0.0051
540.5038	1.3889	500	586.3414	0.0711	0.0041
540.5038	1.6667	600	575.1269	0.0709	0.0044
540.5038	1.9444	700	567.1767	0.0660	0.0034
540.5038	2.2222	800	561.4174	0.0709	0.0039
540.5038	2.5	900	554.0192	0.0750	0.0044
428.2781	2.7778	1000	548.8307	0.0759	0.0075
428.2781	3.0556	1100	544.0135	0.0761	0.0077
428.2781	3.3333	1200	538.4531	0.0736	0.0048
428.2781	3.6111	1300	530.2228	0.0629	0.0073
428.2781	3.8889	1400	525.1834	0.0601	0.0052
372.0535	4.1667	1500	516.5657	0.0513	0.0048
372.0535	4.4444	1600	513.2379	0.0617	0.0091
372.0535	4.7222	1700	505.9874	0.0722	0.0073
372.0535	5.0	1800	497.9989	0.0610	0.0097
372.0535	5.2778	1900	495.0910	0.0616	0.0113
329.5571	5.5556	2000	491.1795	0.0638	0.0102
329.5571	5.8333	2100	486.6686	0.0745	0.0130
329.5571	6.1111	2200	481.7049	0.0653	0.0114
329.5571	6.3889	2300	475.6475	0.0650	0.0121
329.5571	6.6667	2400	470.6474	0.0615	0.0121
298.0146	6.9444	2500	469.8773	0.0749	0.0146
298.0146	7.2222	2600	462.3435	0.0728	0.0152
298.0146	7.5	2700	459.8708	0.0680	0.0149
298.0146	7.7778	2800	456.7115	0.0649	0.0151
298.0146	8.0556	2900	451.0399	0.0691	0.0176
272.1015	8.3333	3000	445.2203	0.0701	0.0178
272.1015	8.6111	3100	439.9684	0.0789	0.0201
272.1015	8.8889	3200	437.1039	0.0705	0.0193
272.1015	9.1667	3300	433.4584	0.0655	0.0184
272.1015	9.4444	3400	428.3048	0.0731	0.0208
249.4726	9.7222	3500	423.4261	0.0639	0.0178
249.4726	10.0	3600	422.1715	0.0661	0.0198
249.4726	10.2778	3700	417.5420	0.0774	0.0234
249.4726	10.5556	3800	416.0034	0.0731	0.0219
249.4726	10.8333	3900	410.5283	0.0822	0.0276
230.0448	11.1111	4000	408.2419	0.0830	0.0278
230.0448	11.3889	4100	403.9637	0.0755	0.0280
230.0448	11.6667	4200	401.9917	0.0684	0.0233
230.0448	11.9444	4300	396.8027	0.0722	0.0263
230.0448	12.2222	4400	394.9419	0.0781	0.0281
212.8543	12.5	4500	390.4717	0.0802	0.0310
212.8543	12.7778	4600	388.3587	0.0744	0.0269
212.8543	13.0556	4700	387.5984	0.0790	0.0279
212.8543	13.3333	4800	383.2326	0.0806	0.0302
212.8543	13.6111	4900	383.7005	0.0736	0.0293
198.2257	13.8889	5000	374.8887	0.0815	0.0330
198.2257	14.1667	5100	373.3404	0.0791	0.0319
198.2257	14.4444	5200	370.0199	0.0747	0.0322
198.2257	14.7222	5300	369.0208	0.0832	0.0361
198.2257	15.0	5400	366.3546	0.0830	0.0341
185.0336	15.2778	5500	363.6714	0.0860	0.0360
185.0336	15.5556	5600	361.8872	0.0843	0.0358
185.0336	15.8333	5700	357.5280	0.0748	0.0328
185.0336	16.1111	5800	358.1822	0.0844	0.0362
185.0336	16.3889	5900	354.5366	0.0848	0.0371
173.9678	16.6667	6000	352.3711	0.0820	0.0366
173.9678	16.9444	6100	350.1420	0.0840	0.0379
173.9678	17.2222	6200	346.6544	0.0862	0.0388
173.9678	17.5	6300	345.2696	0.0854	0.0376
173.9678	17.7778	6400	343.5911	0.0852	0.0379
164.343	18.0556	6500	342.2870	0.0925	0.0399
164.343	18.3333	6600	340.4440	0.0812	0.0384
164.343	18.6111	6700	337.7889	0.0839	0.0391
164.343	18.8889	6800	337.9293	0.0888	0.0387
164.343	19.1667	6900	337.0860	0.0951	0.0432
156.0394	19.4444	7000	338.8325	0.0908	0.0380
156.0394	19.7222	7100	334.3348	0.0913	0.0400
156.0394	20.0	7200	330.6484	0.0863	0.0404
156.0394	20.2778	7300	329.6674	0.0852	0.0419
156.0394	20.5556	7400	326.3730	0.0855	0.0404
149.0754	20.8333	7500	325.8025	0.0820	0.0387
149.0754	21.1111	7600	325.7752	0.0857	0.0421
149.0754	21.3889	7700	325.3289	0.0911	0.0401
149.0754	21.6667	7800	322.8049	0.0869	0.0424
149.0754	21.9444	7900	322.9578	0.0788	0.0385
143.0181	22.2222	8000	322.0035	0.0900	0.0414
143.0181	22.5	8100	320.1973	0.0922	0.0428
143.0181	22.7778	8200	318.8786	0.0887	0.0428
143.0181	23.0556	8300	318.5295	0.0928	0.0429
143.0181	23.3333	8400	316.3871	0.0894	0.0413
138.3268	23.6111	8500	317.5307	0.0890	0.0430
138.3268	23.8889	8600	317.3522	0.0902	0.0432
138.3268	24.1667	8700	316.0612	0.0869	0.0408
138.3268	24.4444	8800	314.3895	0.0875	0.0428
138.3268	24.7222	8900	311.4596	0.0913	0.0437
134.1338	25.0	9000	312.0233	0.0887	0.0416
134.1338	25.2778	9100	310.7538	0.0897	0.0439
134.1338	25.5556	9200	311.2314	0.0911	0.0435
134.1338	25.8333	9300	311.0068	0.0900	0.0414
134.1338	26.1111	9400	309.6226	0.0927	0.0438
131.2272	26.3889	9500	308.3279	0.0872	0.0423
131.2272	26.6667	9600	308.4141	0.0868	0.0428
131.2272	26.9444	9700	308.6096	0.0907	0.0433
131.2272	27.2222	9800	308.6696	0.0913	0.0421
131.2272	27.5	9900	308.0463	0.0907	0.0441
128.9031	27.7778	10000	307.7925	0.0868	0.0431
128.9031	28.0556	10100	306.5993	0.0889	0.0427
128.9031	28.3333	10200	306.2005	0.0932	0.0446
128.9031	28.6111	10300	307.4815	0.0881	0.0432
128.9031	28.8889	10400	307.0386	0.0914	0.0443
127.4595	29.1667	10500	306.5276	0.0901	0.0435
127.4595	29.4444	10600	306.6310	0.0897	0.0435
127.4595	29.7222	10700	306.8422	0.0905	0.0435
127.4595	30.0	10800	306.4669	0.0905	0.0439

Framework versions

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