ht-stmini-cls-v6_ftis_noPretrain-msm-80pos

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2605
• Accuracy: 0.8904
• Macro F1: 0.7012

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 4
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 6733
• training_steps: 134674

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Macro F1
73.8159	0.0013	174	45.3995	0.0656	0.0286
37.5531	1.0013	348	25.1655	0.1618	0.0513
23.72	2.0013	522	17.9662	0.3776	0.0771
17.4471	3.0013	696	13.2464	0.4452	0.0983
11.5736	4.0013	870	9.5088	0.4901	0.1211
8.7443	5.0013	1044	7.3522	0.5245	0.1280
6.586	6.0012	1218	5.5780	0.5222	0.1280
5.6125	7.0012	1392	5.0756	0.5475	0.1336
4.6907	8.0012	1566	4.4535	0.5707	0.1376
4.1955	9.0012	1740	4.0657	0.5854	0.1424
3.9359	10.0012	1914	3.7584	0.5821	0.1430
3.64	11.0012	2088	3.4417	0.5939	0.1461
3.5137	12.0012	2262	3.4068	0.6018	0.1494
3.4638	13.0012	2436	3.3635	0.6036	0.1553
3.1564	14.0012	2610	3.1774	0.6152	0.1516
3.1857	15.0012	2784	3.0813	0.6134	0.1494
2.9395	16.0012	2958	3.1152	0.5992	0.1513
2.8653	17.0012	3132	3.1513	0.5816	0.1467
2.9153	18.0012	3306	2.8326	0.6368	0.1644
2.8393	19.0012	3480	3.0046	0.6101	0.1546
2.7325	20.0011	3654	2.8606	0.6335	0.1754
2.6792	21.0011	3828	2.9711	0.6021	0.1720
2.6199	22.0011	4002	2.8676	0.6311	0.1798
2.5997	23.0011	4176	2.7022	0.6756	0.1979
2.3936	24.0011	4350	2.7604	0.6347	0.1969
2.4254	25.0011	4524	2.6484	0.6714	0.1964
2.2806	26.0011	4698	2.5601	0.6803	0.2177
2.2157	27.0011	4872	2.4793	0.7045	0.2442
2.1541	28.0011	5046	2.5877	0.6823	0.2468
2.1663	29.0011	5220	2.4217	0.6905	0.2713
2.0744	30.0011	5394	2.3407	0.7216	0.2804
2.0074	31.0011	5568	2.3610	0.7163	0.2866
1.9823	32.0011	5742	2.3197	0.7046	0.3021
1.868	33.0010	5916	2.5295	0.7091	0.2808
1.8692	34.0010	6090	2.6165	0.6713	0.2718
1.9232	35.0010	6264	2.1315	0.7410	0.3690
1.8276	36.0010	6438	2.1101	0.7366	0.3463
1.6996	37.0010	6612	2.1115	0.7292	0.3566
1.81	38.0010	6786	2.1167	0.7415	0.3794
1.7055	39.0010	6960	2.2200	0.7420	0.3862
1.5763	40.0010	7134	1.9617	0.7591	0.4129
1.5598	41.0010	7308	2.1212	0.7566	0.3809
1.5548	42.0010	7482	2.0319	0.7720	0.4159
1.6272	43.0010	7656	1.8861	0.7675	0.4212
1.4524	44.0010	7830	1.9266	0.7619	0.4208
1.4248	45.0010	8004	2.0008	0.7665	0.4362
1.3838	46.0010	8178	1.9488	0.7577	0.4330
1.3397	47.0009	8352	1.8366	0.7871	0.4667
1.3212	48.0009	8526	1.9344	0.7761	0.4492
1.256	49.0009	8700	1.9458	0.7625	0.4621
1.2357	50.0009	8874	1.9950	0.7679	0.4412
1.3454	51.0009	9048	1.8599	0.7880	0.4714
1.239	52.0009	9222	1.7790	0.7945	0.4809
1.2405	53.0009	9396	1.9177	0.7886	0.4729
1.1847	54.0009	9570	1.7395	0.7971	0.4877
1.0741	55.0009	9744	1.7458	0.7945	0.4821
1.0478	56.0009	9918	1.7530	0.7984	0.5031
1.0162	57.0009	10092	1.6300	0.8096	0.5140
1.0867	58.0009	10266	1.8345	0.7928	0.5056
1.0302	59.0009	10440	1.7416	0.7980	0.4959
1.054	60.0008	10614	1.6987	0.8033	0.5133
0.9682	61.0008	10788	1.7021	0.8075	0.5135
0.974	62.0008	10962	1.9826	0.7884	0.4771
1.0976	63.0008	11136	1.8826	0.7965	0.4894
0.9831	64.0008	11310	1.8587	0.7962	0.5128
0.9037	65.0008	11484	1.7401	0.8059	0.5263
0.8688	66.0008	11658	1.6962	0.8082	0.5101
0.8852	67.0008	11832	1.6819	0.8151	0.5343
0.8469	68.0008	12006	1.6921	0.8154	0.5225
0.8645	69.0008	12180	1.6661	0.8163	0.5264
0.8264	70.0008	12354	1.6807	0.8148	0.5311
0.8552	71.0008	12528	1.6552	0.8213	0.5330
0.9329	72.0008	12702	1.6248	0.8148	0.5197
0.8395	73.0007	12876	1.7094	0.8217	0.5326
0.8221	74.0007	13050	1.7074	0.8263	0.5499
0.7703	75.0007	13224	1.6904	0.8264	0.5467
0.7861	76.0007	13398	1.6314	0.8282	0.5514
0.8042	77.0007	13572	1.6830	0.8248	0.5460
0.8099	78.0007	13746	1.8110	0.8171	0.5308
0.739	79.0007	13920	1.6330	0.8195	0.5432
0.8146	80.0007	14094	1.6790	0.8161	0.5353
0.7428	81.0007	14268	1.5951	0.8291	0.5535
0.7585	82.0007	14442	1.6879	0.8348	0.5614
0.7248	83.0007	14616	1.6282	0.8295	0.5595
0.7213	84.0007	14790	1.5450	0.8353	0.5672
0.6816	85.0007	14964	1.5704	0.8334	0.5710
0.6988	86.0007	15138	1.6659	0.8273	0.5611
0.6981	87.0006	15312	1.6111	0.8362	0.5667
0.6545	88.0006	15486	1.5741	0.8344	0.5751
0.6597	89.0006	15660	1.6150	0.8345	0.5695
0.6431	90.0006	15834	1.5277	0.8344	0.5758
0.6524	91.0006	16008	1.5278	0.8407	0.5840
0.6746	92.0006	16182	1.6709	0.8341	0.5773
0.6313	93.0006	16356	1.5784	0.8427	0.5884
0.6199	94.0006	16530	1.5905	0.8418	0.5786
0.608	95.0006	16704	1.4833	0.8439	0.5837
0.6471	96.0006	16878	1.5221	0.8441	0.5917
0.6516	97.0006	17052	1.5060	0.8405	0.5772
0.6386	98.0006	17226	1.6709	0.8387	0.5819
0.6183	99.0006	17400	1.4718	0.8401	0.5789
0.6194	100.0005	17574	1.4719	0.8506	0.6074
0.6311	101.0005	17748	1.5237	0.8509	0.6060
0.5862	102.0005	17922	1.4453	0.8547	0.6097
0.649	103.0005	18096	1.5579	0.8456	0.5993
0.5825	104.0005	18270	1.6311	0.8478	0.5963
0.5912	105.0005	18444	1.6532	0.8449	0.5929
0.5848	106.0005	18618	1.4357	0.8523	0.6010
0.6126	107.0005	18792	1.4119	0.8575	0.6127
0.5385	108.0005	18966	1.4446	0.8555	0.6142
0.5465	109.0005	19140	1.6441	0.8461	0.5965
0.6138	110.0005	19314	1.5808	0.8485	0.6039
0.5527	111.0005	19488	1.4936	0.8510	0.6099
0.5474	112.0005	19662	1.4943	0.8545	0.6166
0.5191	113.0005	19836	1.4353	0.8549	0.6198
0.5216	114.0004	20010	1.4123	0.8574	0.6182
0.5278	115.0004	20184	1.4450	0.8528	0.6181
0.5489	116.0004	20358	1.4651	0.8590	0.6243
0.551	117.0004	20532	1.4128	0.8587	0.6216
0.5084	118.0004	20706	1.5829	0.8546	0.6170
0.4925	119.0004	20880	1.4310	0.8591	0.6245
0.5197	120.0004	21054	1.4539	0.8665	0.6323
0.4741	121.0004	21228	1.4279	0.8627	0.6356
0.5152	122.0004	21402	1.7166	0.8602	0.6197
0.4913	123.0004	21576	1.3961	0.8655	0.6322
0.4872	124.0004	21750	1.4094	0.8565	0.6350
0.4781	125.0004	21924	1.3850	0.8602	0.6308
0.4993	126.0004	22098	1.4830	0.8607	0.6207
0.5007	127.0003	22272	1.4839	0.8583	0.6262
0.4722	128.0003	22446	1.4310	0.8697	0.6420
0.5139	129.0003	22620	1.3813	0.8665	0.6453
0.502	130.0003	22794	1.3679	0.8728	0.6499
0.4801	131.0003	22968	1.4264	0.8688	0.6420
0.4702	132.0003	23142	1.4473	0.8670	0.6438
0.4667	133.0003	23316	1.4691	0.8685	0.6403
0.4673	134.0003	23490	1.3438	0.8744	0.6539
0.4577	135.0003	23664	1.4150	0.8686	0.6468
0.4483	136.0003	23838	1.2918	0.8738	0.6497
0.4688	137.0003	24012	1.3487	0.8675	0.6469
0.4467	138.0003	24186	1.4824	0.8632	0.6386
0.4718	139.0003	24360	1.3348	0.8694	0.6493
0.4716	140.0003	24534	1.4506	0.8711	0.6423
0.4432	141.0002	24708	1.3371	0.8657	0.6440
0.4304	142.0002	24882	1.4240	0.8701	0.6488
0.4821	143.0002	25056	1.4996	0.8633	0.6376
0.4707	144.0002	25230	1.4272	0.8732	0.6582
0.4569	145.0002	25404	1.4972	0.8532	0.6420
0.4715	146.0002	25578	1.3520	0.8659	0.6471
0.464	147.0002	25752	1.4628	0.8687	0.6453
0.4402	148.0002	25926	1.3474	0.8725	0.6508
0.4176	149.0002	26100	1.4306	0.8781	0.6613
0.3966	150.0002	26274	1.3421	0.8781	0.6605
0.4043	151.0002	26448	1.3607	0.8772	0.6597
0.4355	152.0002	26622	1.2996	0.8772	0.6592
0.4058	153.0002	26796	1.3772	0.8711	0.6478
0.4232	154.0001	26970	1.2894	0.8758	0.6650
0.3934	155.0001	27144	1.4422	0.8718	0.6549
0.4018	156.0001	27318	1.3746	0.8740	0.6598
0.4182	157.0001	27492	1.3915	0.8759	0.6641
0.3839	158.0001	27666	1.4101	0.8783	0.6679
0.3963	159.0001	27840	1.4369	0.8746	0.6601
0.3847	160.0001	28014	1.4886	0.8684	0.6518
0.3832	161.0001	28188	1.2403	0.8838	0.6762
0.3905	162.0001	28362	1.3156	0.8816	0.6725
0.3963	163.0001	28536	1.3896	0.8670	0.6584
0.4108	164.0001	28710	1.3996	0.8771	0.6658
0.4154	165.0001	28884	1.4120	0.8757	0.6657
0.3753	166.0001	29058	1.2419	0.8796	0.6716
0.3853	167.0001	29232	1.6330	0.8768	0.6715
0.3759	168.0000	29406	1.3326	0.8795	0.6785
0.3746	169.0000	29580	1.3901	0.8746	0.6692
0.38	170.0000	29754	1.4251	0.8788	0.6756
0.3767	171.0000	29928	1.3926	0.8720	0.6812
0.4058	172.0000	30102	1.3462	0.8760	0.6719
0.364	173.0000	30276	1.3386	0.8809	0.6738
0.3742	173.0013	30450	1.3045	0.8836	0.6913
0.3731	174.0013	30624	1.4262	0.8743	0.6683
0.3583	175.0013	30798	1.3757	0.8778	0.6742
0.3865	176.0013	30972	1.4321	0.8782	0.6725
0.3681	177.0013	31146	1.3325	0.8816	0.6801
0.3626	178.0013	31320	1.2028	0.8866	0.6876
0.3662	179.0013	31494	1.3165	0.8848	0.6892
0.3513	180.0012	31668	1.3460	0.8852	0.6822
0.3885	181.0012	31842	1.5075	0.8806	0.6818
0.3436	182.0012	32016	1.3440	0.8870	0.6835
0.3651	183.0012	32190	1.3339	0.8850	0.6896
0.3457	184.0012	32364	1.3043	0.8874	0.6925
0.3474	185.0012	32538	1.3221	0.8872	0.6908
0.3685	186.0012	32712	1.3012	0.8857	0.6952
0.3581	187.0012	32886	1.2899	0.8893	0.7009
0.3431	188.0012	33060	1.5158	0.8758	0.6721
0.3699	189.0012	33234	1.3135	0.8846	0.6955
0.3633	190.0012	33408	1.3714	0.8757	0.6790
0.3627	191.0012	33582	1.3396	0.8815	0.6769
0.3404	192.0012	33756	1.2124	0.8861	0.6941
0.3385	193.0012	33930	1.4343	0.8794	0.6856
0.3459	194.0011	34104	1.4034	0.8855	0.6881
0.3717	195.0011	34278	1.4770	0.8765	0.6753
0.3616	196.0011	34452	1.2764	0.8850	0.6976
0.3418	197.0011	34626	1.3811	0.8843	0.6876
0.3346	198.0011	34800	1.2929	0.8844	0.6894
0.336	199.0011	34974	1.3234	0.8880	0.6966
0.3413	200.0011	35148	1.1831	0.8898	0.6979
0.3324	201.0011	35322	1.2362	0.8875	0.6978
0.3435	202.0011	35496	1.2785	0.8885	0.6932
0.3709	203.0011	35670	1.3519	0.8877	0.6975
0.3483	204.0011	35844	1.3038	0.8829	0.6969
0.3409	205.0011	36018	1.4166	0.8816	0.6827
0.3353	206.0011	36192	1.2447	0.8893	0.6935
0.35	207.0010	36366	1.2934	0.8910	0.7006

Framework versions

• Transformers 4.46.0
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.20.1