ht-stmini-cls-v6_ftis_noPretrain-msm-pos

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

• Loss: 1.8172
• Accuracy: 0.8925
• Macro F1: 0.7511

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: 134675

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Macro F1
59.3464	0.0013	169	38.9204	0.0790	0.0371
25.3687	1.0012	338	95.5343	0.1876	0.0584
7.8147	2.0012	507	195.3371	0.4964	0.1235
6.8891	3.0012	676	191.6926	0.5398	0.1325
6.2177	4.0012	845	216.1844	0.5697	0.1383
5.4678	5.0012	1014	189.3503	0.5881	0.1441
4.9594	6.0012	1183	137.5813	0.5951	0.1504
4.1078	7.0012	1352	94.4770	0.6093	0.1529
3.5533	8.0012	1521	64.6198	0.6143	0.1643
3.2608	9.0012	1690	45.8943	0.6192	0.1693
2.9641	10.0012	1859	39.1380	0.6195	0.1724
2.7522	11.0012	2028	33.9856	0.6396	0.1836
2.7176	12.0012	2197	29.4991	0.6150	0.1799
2.5893	13.0012	2366	22.8487	0.6356	0.2027
2.457	14.0012	2535	19.0762	0.6483	0.2075
2.4692	15.0011	2704	16.5434	0.6498	0.2202
2.309	16.0011	2873	14.8761	0.6640	0.2347
2.3723	17.0011	3042	12.4455	0.6624	0.2501
2.2035	18.0011	3211	12.1223	0.6490	0.2582
2.1096	19.0011	3380	12.5485	0.6854	0.2777
1.9845	20.0011	3549	10.0986	0.6758	0.2975
1.9108	21.0011	3718	10.0700	0.6911	0.3001
1.9174	22.0011	3887	6.9076	0.7244	0.3483
1.8328	23.0011	4056	9.2308	0.7231	0.3387
1.6883	24.0011	4225	7.6798	0.7277	0.3699
1.6764	25.0011	4394	7.3750	0.7291	0.3764
1.5494	26.0011	4563	7.9343	0.7419	0.3963
1.4642	27.0011	4732	10.0410	0.7604	0.4199
1.3226	28.0010	4901	7.8654	0.7586	0.4226
1.3147	29.0010	5070	9.4925	0.7783	0.4468
1.2191	30.0010	5239	9.3531	0.7751	0.4472
1.2129	31.0010	5408	9.6055	0.7794	0.4657
1.1316	32.0010	5577	9.9082	0.7883	0.4837
1.1428	33.0010	5746	10.4082	0.7953	0.4963
1.086	34.0010	5915	10.5887	0.8013	0.5049
0.9935	35.0010	6084	9.5130	0.7909	0.4851
0.976	36.0010	6253	11.5568	0.7926	0.5008
0.9635	37.0010	6422	14.4823	0.8069	0.5124
0.8799	38.0010	6591	12.5622	0.7883	0.5165
0.9134	39.0010	6760	14.6331	0.8065	0.5279
0.7942	40.0010	6929	14.0959	0.8103	0.5368
0.7707	41.0010	7098	14.0539	0.7993	0.5011
0.7535	42.0009	7267	12.6058	0.8070	0.5297
0.6551	43.0009	7436	16.7901	0.8179	0.5502
0.6933	44.0009	7605	15.4039	0.8189	0.5602
0.6253	45.0009	7774	17.1676	0.8196	0.5664
0.5847	46.0009	7943	17.8274	0.8237	0.5642
0.5747	47.0009	8112	18.4719	0.8216	0.5661
0.5041	48.0009	8281	17.1367	0.8210	0.5588
0.5139	49.0009	8450	16.5208	0.8356	0.5884
0.4726	50.0009	8619	21.7155	0.8230	0.5779
0.4511	51.0009	8788	22.6982	0.8353	0.5928
0.4668	52.0009	8957	19.2348	0.8347	0.5851
0.3898	53.0009	9126	18.7129	0.8342	0.5923
0.3695	54.0009	9295	18.8422	0.8399	0.5990
0.3801	55.0008	9464	18.5114	0.8413	0.6028
0.3443	56.0008	9633	16.4898	0.8432	0.6050
0.3616	57.0008	9802	18.9793	0.8486	0.6226
0.3354	58.0008	9971	17.3367	0.8461	0.6132
0.3294	59.0008	10140	18.4233	0.8450	0.6191
0.3136	60.0008	10309	18.7376	0.8482	0.6177
0.2854	61.0008	10478	15.7951	0.8552	0.6289
0.2838	62.0008	10647	17.5900	0.8503	0.6238
0.2536	63.0008	10816	16.5078	0.8525	0.6307
0.2617	64.0008	10985	17.8552	0.8454	0.6350
0.2461	65.0008	11154	16.2489	0.8569	0.6432
0.237	66.0008	11323	12.4754	0.8532	0.6423
0.2176	67.0008	11492	10.9631	0.8579	0.6467
0.2119	68.0007	11661	13.9939	0.8601	0.6474
0.2068	69.0007	11830	12.4334	0.8532	0.6404
0.1956	70.0007	11999	14.0338	0.8593	0.6504
0.1878	71.0007	12168	11.7233	0.8602	0.6506
0.1875	72.0007	12337	10.3088	0.8642	0.6645
0.2062	73.0007	12506	9.0553	0.8611	0.6599
0.1822	74.0007	12675	7.1633	0.8624	0.6625
0.171	75.0007	12844	8.1621	0.8593	0.6589
0.165	76.0007	13013	7.7144	0.8647	0.6712
0.1602	77.0007	13182	8.1336	0.8627	0.6610
0.1586	78.0007	13351	6.4341	0.8619	0.6650
0.154	79.0007	13520	6.1606	0.8663	0.6677
0.1455	80.0007	13689	6.5879	0.8625	0.6630
0.1442	81.0007	13858	6.1570	0.8701	0.6782
0.1419	82.0006	14027	6.2919	0.8675	0.6719
0.1372	83.0006	14196	5.1177	0.8662	0.6750
0.1181	84.0006	14365	4.8350	0.8666	0.6795
0.1252	85.0006	14534	4.3481	0.8711	0.6811
0.1269	86.0006	14703	4.7319	0.8656	0.6819
0.123	87.0006	14872	4.7458	0.8714	0.6839
0.1278	88.0006	15041	3.9598	0.8708	0.6858
0.1122	89.0006	15210	3.5908	0.8754	0.6937
0.117	90.0006	15379	3.8412	0.8721	0.6865
0.1141	91.0006	15548	3.8014	0.8730	0.6889
0.1064	92.0006	15717	3.6323	0.8726	0.6889
0.1106	93.0006	15886	3.5584	0.8734	0.6908
0.1019	94.0006	16055	3.3942	0.8738	0.6897
0.0961	95.0005	16224	3.2197	0.8749	0.6961
0.0992	96.0005	16393	3.3110	0.8706	0.6976
0.1046	97.0005	16562	2.9395	0.8757	0.6981
0.0967	98.0005	16731	3.0819	0.8715	0.6927
0.0852	99.0005	16900	2.7328	0.8725	0.6929
0.0909	100.0005	17069	2.6428	0.8763	0.7052
0.0884	101.0005	17238	2.6717	0.8754	0.7019
0.0877	102.0005	17407	2.6857	0.8721	0.7029
0.0787	103.0005	17576	2.5938	0.8768	0.7032
0.0849	104.0005	17745	2.5608	0.8771	0.7090
0.08	105.0005	17914	2.5091	0.8756	0.6984
0.074	106.0005	18083	2.5128	0.8803	0.7063
0.0714	107.0005	18252	2.4161	0.8773	0.7068
0.0771	108.0005	18421	2.2031	0.8794	0.7094
0.0895	109.0004	18590	2.4464	0.8804	0.7163
0.0728	110.0004	18759	2.3420	0.8759	0.7125
0.0971	111.0004	18928	2.3484	0.8781	0.7017
0.0752	112.0004	19097	2.2271	0.8790	0.7178
0.067	113.0004	19266	2.4369	0.8765	0.7120
0.0741	114.0004	19435	2.2862	0.8777	0.7139
0.0738	115.0004	19604	2.2621	0.8785	0.7091
0.062	116.0004	19773	2.3678	0.8787	0.7111
0.0747	117.0004	19942	2.4907	0.8745	0.6961
0.0653	118.0004	20111	2.1874	0.8827	0.7161
0.0658	119.0004	20280	2.0325	0.8838	0.7217
0.0652	120.0004	20449	2.1266	0.8820	0.7156
0.0597	121.0004	20618	2.1654	0.8801	0.7149
0.0757	122.0003	20787	1.9499	0.8866	0.7287
0.0593	123.0003	20956	2.2619	0.8791	0.7118
0.0576	124.0003	21125	2.1054	0.8829	0.7184
0.0616	125.0003	21294	2.1809	0.8796	0.7182
0.0601	126.0003	21463	1.8629	0.8872	0.7277
0.0548	127.0003	21632	2.1090	0.8821	0.7166
0.0572	128.0003	21801	2.2149	0.8784	0.7153
0.0561	129.0003	21970	1.9411	0.8856	0.7303
0.0554	130.0003	22139	2.0311	0.8813	0.7183
0.0509	131.0003	22308	2.1824	0.8839	0.7188
0.0616	132.0003	22477	2.1547	0.8830	0.7189
0.0511	133.0003	22646	2.0628	0.8816	0.7195
0.0491	134.0003	22815	1.9718	0.8859	0.7266
0.0564	135.0003	22984	2.0992	0.8835	0.7223
0.0477	136.0002	23153	1.9783	0.8828	0.7216
0.049	137.0002	23322	2.1633	0.8851	0.7205
0.0448	138.0002	23491	2.0866	0.8815	0.7248
0.0508	139.0002	23660	2.2082	0.8861	0.7229
0.0561	140.0002	23829	2.0297	0.8831	0.7263
0.0476	141.0002	23998	2.1306	0.8822	0.7247
0.0562	142.0002	24167	2.2599	0.8819	0.7182
0.0433	143.0002	24336	2.0900	0.8846	0.7249
0.0454	144.0002	24505	2.0576	0.8852	0.7280
0.0439	145.0002	24674	1.9070	0.8861	0.7354
0.0427	146.0002	24843	1.9406	0.8859	0.7255
0.0435	147.0002	25012	1.8753	0.8858	0.7350
0.0386	148.0002	25181	2.0958	0.8831	0.7261
0.0421	149.0001	25350	2.0531	0.8840	0.7237
0.0418	150.0001	25519	2.0721	0.8848	0.7228
0.0392	151.0001	25688	2.2118	0.8828	0.7232
0.0431	152.0001	25857	2.3609	0.8840	0.7200
0.0449	153.0001	26026	1.9814	0.8851	0.7302
0.0395	154.0001	26195	1.9841	0.8889	0.7342
0.0428	155.0001	26364	2.1003	0.8851	0.7314
0.0378	156.0001	26533	2.0011	0.8899	0.7357
0.0417	157.0001	26702	1.9045	0.8885	0.7313
0.0369	158.0001	26871	2.1463	0.8817	0.7237
0.0381	159.0001	27040	1.8939	0.8891	0.7359
0.039	160.0001	27209	2.1339	0.8860	0.7280
0.0409	161.0001	27378	2.0200	0.8849	0.7332
0.0357	162.0001	27547	1.9856	0.8894	0.7385
0.0388	163.0000	27716	1.8941	0.8889	0.7352
0.0405	164.0000	27885	2.1008	0.8845	0.7219
0.0392	165.0000	28054	2.0878	0.8837	0.7295
0.0361	166.0000	28223	1.8980	0.8880	0.7381
0.0353	167.0000	28392	1.9293	0.8882	0.7324
0.0344	168.0000	28561	1.8784	0.8889	0.7405
0.0327	168.0013	28730	1.8787	0.8903	0.7377
0.032	169.0013	28899	2.2492	0.8857	0.7300
0.0348	170.0012	29068	2.1504	0.8874	0.7359
0.0377	171.0012	29237	2.1245	0.8897	0.7373
0.0343	172.0012	29406	1.9574	0.8892	0.7370
0.0337	173.0012	29575	1.9612	0.8885	0.7381
0.0375	174.0012	29744	2.0124	0.8905	0.7427
0.0314	175.0012	29913	2.0747	0.8893	0.7352
0.0328	176.0012	30082	2.0896	0.8876	0.7319
0.0313	177.0012	30251	1.9872	0.8885	0.7355
0.0325	178.0012	30420	2.3101	0.8868	0.7323
0.0323	179.0012	30589	1.8263	0.8925	0.7511
0.0368	180.0012	30758	1.9961	0.8904	0.7403
0.0325	181.0012	30927	2.0162	0.8897	0.7433
0.0318	182.0012	31096	1.9951	0.8882	0.7396
0.033	183.0012	31265	1.7084	0.8951	0.7484
0.0308	184.0011	31434	2.0011	0.8884	0.7386
0.0302	185.0011	31603	1.9691	0.8909	0.7389
0.0312	186.0011	31772	2.0769	0.8913	0.7410
0.0334	187.0011	31941	2.1664	0.8862	0.7300
0.0295	188.0011	32110	2.1281	0.8898	0.7387
0.0287	189.0011	32279	2.3105	0.8861	0.7256
0.033	190.0011	32448	1.9435	0.8925	0.7389
0.029	191.0011	32617	2.0848	0.8894	0.7417
0.032	192.0011	32786	2.0616	0.8869	0.7329
0.029	193.0011	32955	2.1727	0.8867	0.7336
0.0254	194.0011	33124	2.1293	0.8911	0.7405
0.0263	195.0011	33293	2.0403	0.8916	0.7358
0.0283	196.0011	33462	2.1992	0.8904	0.7407
0.0308	197.0010	33631	2.1202	0.8900	0.7379
0.0274	198.0010	33800	2.0473	0.8911	0.7401
0.0255	199.0010	33969	2.1266	0.8914	0.7441

Framework versions

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