ht-stmini-cls-v6_ftis_noPretrain-gtsp-m1drp0.5trp0.5

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

• Loss: 7.1405
• Accuracy: 0.8861
• Macro F1: 0.7242

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
46.1739	0.0013	169	54.5936	0.0643	0.0263
20.3382	1.0012	338	119.0171	0.2263	0.0639
7.5451	2.0012	507	196.3073	0.4877	0.1232
6.4507	3.0012	676	232.9836	0.5399	0.1330
6.1129	4.0012	845	196.8743	0.5636	0.1387
5.1204	5.0012	1014	152.6818	0.5846	0.1470
4.6213	6.0012	1183	92.0791	0.6067	0.1570
4.0126	7.0012	1352	71.1870	0.6149	0.1629
3.5806	8.0012	1521	50.8923	0.6144	0.1630
3.3992	9.0012	1690	42.1669	0.6199	0.1712
3.1309	10.0012	1859	32.4374	0.6282	0.1759
2.9659	11.0012	2028	29.6463	0.6272	0.1763
2.8683	12.0012	2197	26.0668	0.6365	0.1933
2.7812	13.0012	2366	23.0157	0.6091	0.1905
2.7425	14.0012	2535	21.0243	0.6446	0.2091
2.6172	15.0011	2704	17.7302	0.6438	0.2232
2.5982	16.0011	2873	15.5038	0.6572	0.2453
2.5605	17.0011	3042	14.4028	0.6640	0.2521
2.5301	18.0011	3211	14.0581	0.6705	0.2627
2.5094	19.0011	3380	12.6933	0.6843	0.2805
2.279	20.0011	3549	12.7472	0.6893	0.3161
2.2173	21.0011	3718	9.8769	0.6950	0.3046
2.1963	22.0011	3887	11.8893	0.7077	0.3428
2.198	23.0011	4056	9.4825	0.7239	0.3676
2.1668	24.0011	4225	9.7164	0.7212	0.3819
2.1268	25.0011	4394	10.6331	0.7240	0.3796
1.9588	26.0011	4563	9.8210	0.7349	0.3834
1.9623	27.0011	4732	9.9962	0.7310	0.4091
1.9763	28.0010	4901	9.9540	0.7505	0.4347
1.9257	29.0010	5070	10.5623	0.7316	0.4262
1.8615	30.0010	5239	10.8115	0.7330	0.4152
1.7739	31.0010	5408	9.5557	0.7539	0.4343
1.7401	32.0010	5577	11.9211	0.7550	0.4510
1.7629	33.0010	5746	10.6967	0.7604	0.4623
1.5682	34.0010	5915	11.4143	0.7473	0.4569
1.6434	35.0010	6084	12.1514	0.7570	0.4817
1.5984	36.0010	6253	12.7246	0.7668	0.4595
1.5916	37.0010	6422	13.0271	0.7673	0.4958
1.4365	38.0010	6591	14.2249	0.7810	0.4872
1.474	39.0010	6760	14.7875	0.7724	0.4964
1.439	40.0010	6929	15.0381	0.7740	0.4937
1.3501	41.0010	7098	13.4024	0.7865	0.5106
1.3817	42.0009	7267	16.2374	0.7803	0.5070
1.3503	43.0009	7436	17.2895	0.7678	0.5064
1.2407	44.0009	7605	18.8522	0.7982	0.5280
1.2298	45.0009	7774	22.5457	0.7926	0.5310
1.199	46.0009	7943	23.9661	0.7969	0.5364
1.1683	47.0009	8112	21.2598	0.7984	0.5427
1.2042	48.0009	8281	23.5666	0.8057	0.5327
1.138	49.0009	8450	23.1386	0.8093	0.5430
1.1506	50.0009	8619	24.9053	0.8102	0.5382
1.0671	51.0009	8788	26.1047	0.8144	0.5545
1.0807	52.0009	8957	22.8301	0.8103	0.5458
1.0567	53.0009	9126	23.2382	0.8192	0.5554
1.0202	54.0009	9295	23.4562	0.8181	0.5539
1.0404	55.0008	9464	22.3486	0.8199	0.5741
1.0291	56.0008	9633	23.9191	0.8171	0.5530
0.9989	57.0008	9802	23.5076	0.8246	0.5564
0.9898	58.0008	9971	23.0975	0.8225	0.5626
1.0024	59.0008	10140	20.3025	0.8171	0.5613
0.9471	60.0008	10309	22.5295	0.8255	0.5657
0.9497	61.0008	10478	21.4868	0.8292	0.5835
0.9352	62.0008	10647	21.2701	0.8318	0.5734
0.9016	63.0008	10816	23.1550	0.8280	0.5742
0.901	64.0008	10985	19.7192	0.8334	0.5856
0.8881	65.0008	11154	21.3281	0.8391	0.5902
0.8847	66.0008	11323	21.6058	0.8310	0.5813
0.8833	67.0008	11492	18.6134	0.8369	0.5883
0.8615	68.0007	11661	20.1302	0.8386	0.5958
0.8676	69.0007	11830	18.9128	0.8338	0.5906
0.8571	70.0007	11999	19.4161	0.8398	0.5945
0.8375	71.0007	12168	17.3181	0.8415	0.6028
0.8295	72.0007	12337	17.6274	0.8447	0.6123
0.8167	73.0007	12506	16.9940	0.8411	0.6070
0.8002	74.0007	12675	15.4095	0.8387	0.6052
0.8253	75.0007	12844	16.7561	0.8433	0.6147
0.7955	76.0007	13013	17.0732	0.8444	0.6172
0.7825	77.0007	13182	15.7842	0.8423	0.6114
0.7873	78.0007	13351	14.9193	0.8463	0.6171
0.7843	79.0007	13520	12.0909	0.8551	0.6335
0.7736	80.0007	13689	15.5886	0.8533	0.6267
0.7732	81.0007	13858	14.8050	0.8481	0.6247
0.7651	82.0006	14027	13.9704	0.8494	0.6312
0.7658	83.0006	14196	12.7090	0.8521	0.6265
0.7744	84.0006	14365	12.8064	0.8540	0.6414
0.759	85.0006	14534	11.2890	0.8539	0.6375
0.7522	86.0006	14703	12.0294	0.8537	0.6357
0.7486	87.0006	14872	11.4561	0.8554	0.6348
0.7466	88.0006	15041	10.2362	0.8571	0.6459
0.7474	89.0006	15210	10.3035	0.8595	0.6440
0.7261	90.0006	15379	11.0865	0.8548	0.6378
0.7516	91.0006	15548	11.3736	0.8557	0.6375
0.731	92.0006	15717	10.4393	0.8540	0.6427
0.7284	93.0006	15886	10.0650	0.8611	0.6552
0.7199	94.0006	16055	9.7093	0.8596	0.6481
0.7113	95.0005	16224	10.9101	0.8563	0.6462
0.719	96.0005	16393	10.2613	0.8609	0.6630
0.7131	97.0005	16562	9.6693	0.8645	0.6649
0.7105	98.0005	16731	10.5206	0.8591	0.6627
0.7058	99.0005	16900	8.4620	0.8652	0.6582
0.7024	100.0005	17069	8.8105	0.8643	0.6570
0.7022	101.0005	17238	9.3463	0.8666	0.6626
0.6979	102.0005	17407	9.0670	0.8632	0.6578
0.699	103.0005	17576	9.1035	0.8644	0.6628
0.7127	104.0005	17745	7.6414	0.8656	0.6662
0.6951	105.0005	17914	9.0160	0.8670	0.6697
0.7026	106.0005	18083	8.6094	0.8647	0.6639
0.6952	107.0005	18252	8.4375	0.8649	0.6636
0.6842	108.0005	18421	7.9563	0.8634	0.6616
0.7037	109.0004	18590	7.6682	0.8645	0.6673
0.6824	110.0004	18759	8.2813	0.8635	0.6665
0.6878	111.0004	18928	7.9821	0.8662	0.6719
0.6832	112.0004	19097	8.8222	0.8642	0.6672
0.68	113.0004	19266	8.4816	0.8711	0.6752
0.6887	114.0004	19435	7.5124	0.8672	0.6743
0.6809	115.0004	19604	7.4288	0.8689	0.6726
0.6816	116.0004	19773	8.2735	0.8738	0.6789
0.6745	117.0004	19942	7.3800	0.8759	0.6833
0.674	118.0004	20111	6.7738	0.8730	0.6851
0.6746	119.0004	20280	7.3061	0.8698	0.6788
0.6773	120.0004	20449	7.0637	0.8752	0.6808
0.6673	121.0004	20618	8.0335	0.8732	0.6857
0.6691	122.0003	20787	7.5896	0.8699	0.6825
0.6717	123.0003	20956	7.5674	0.8658	0.6750
0.6642	124.0003	21125	7.0017	0.8713	0.6840
0.6708	125.0003	21294	6.9829	0.8648	0.6769
0.6756	126.0003	21463	7.8395	0.8716	0.6828
0.6525	127.0003	21632	6.1401	0.8730	0.6895
0.6621	128.0003	21801	7.0304	0.8723	0.6846
0.6494	129.0003	21970	6.6580	0.8713	0.6823
0.6636	130.0003	22139	6.0764	0.8736	0.6876
0.6599	131.0003	22308	6.0239	0.8689	0.6772
0.6579	132.0003	22477	6.0557	0.8733	0.6849
0.6639	133.0003	22646	6.0861	0.8771	0.6930
0.66	134.0003	22815	7.9144	0.8744	0.6928
0.6561	135.0003	22984	6.9709	0.8695	0.6841
0.651	136.0002	23153	6.3749	0.8795	0.6951
0.6577	137.0002	23322	7.3569	0.8730	0.6885
0.657	138.0002	23491	6.4300	0.8738	0.6889
0.6529	139.0002	23660	7.2328	0.8727	0.6863
0.6503	140.0002	23829	6.2251	0.8779	0.6937
0.6493	141.0002	23998	5.4486	0.8762	0.6954
0.6506	142.0002	24167	7.0720	0.8732	0.6916
0.6497	143.0002	24336	5.9793	0.8778	0.6941
0.652	144.0002	24505	6.6768	0.8752	0.6950
0.6453	145.0002	24674	5.6576	0.8766	0.6960
0.6477	146.0002	24843	6.5615	0.8771	0.6971
0.6405	147.0002	25012	5.6985	0.8771	0.6953
0.6458	148.0002	25181	6.7291	0.8758	0.6947
0.6423	149.0001	25350	7.1220	0.8741	0.6936
0.6426	150.0001	25519	6.8760	0.8770	0.6990
0.6411	151.0001	25688	6.8829	0.8758	0.6968
0.6521	152.0001	25857	6.0781	0.8751	0.6922
0.6342	153.0001	26026	7.5541	0.8787	0.7024
0.6385	154.0001	26195	6.5542	0.8751	0.7010
0.6341	155.0001	26364	6.9202	0.8777	0.7025
0.6416	156.0001	26533	6.2821	0.8750	0.6934
0.6417	157.0001	26702	5.9760	0.8787	0.7036
0.6375	158.0001	26871	6.9215	0.8749	0.6941
0.6396	159.0001	27040	6.9586	0.8755	0.6988
0.6402	160.0001	27209	7.4873	0.8775	0.6967
0.6384	161.0001	27378	6.6340	0.8765	0.6946
0.6417	162.0001	27547	6.9707	0.8791	0.6991
0.6347	163.0000	27716	6.7165	0.8759	0.6989
0.636	164.0000	27885	7.1870	0.8758	0.7025
0.6343	165.0000	28054	6.4669	0.8839	0.7085
0.6304	166.0000	28223	6.3772	0.8783	0.6999
0.6251	167.0000	28392	6.7920	0.8760	0.6977
0.6365	168.0000	28561	6.4200	0.8780	0.7018
0.6278	168.0013	28730	6.4624	0.8704	0.6910
0.6291	169.0013	28899	5.7714	0.8754	0.6975
0.6262	170.0012	29068	6.3103	0.8738	0.6979
0.6233	171.0012	29237	6.5736	0.8738	0.6973
0.6238	172.0012	29406	6.3100	0.8805	0.7077
0.6267	173.0012	29575	6.5377	0.8798	0.7111
0.6237	174.0012	29744	6.6138	0.8822	0.7099
0.6243	175.0012	29913	7.3389	0.8809	0.7056
0.6268	176.0012	30082	7.9599	0.8803	0.7083
0.6331	177.0012	30251	6.3465	0.8812	0.7077
0.6275	178.0012	30420	6.0927	0.8818	0.7153
0.625	179.0012	30589	6.7591	0.8814	0.7071
0.6299	180.0012	30758	6.3475	0.8800	0.7095
0.6188	181.0012	30927	7.6586	0.8735	0.6939
0.6215	182.0012	31096	6.3617	0.8842	0.7120
0.6228	183.0012	31265	6.9955	0.8805	0.7172
0.6171	184.0011	31434	6.6193	0.8811	0.7154
0.6194	185.0011	31603	6.4875	0.8814	0.7128
0.6187	186.0011	31772	5.4354	0.8824	0.7122
0.619	187.0011	31941	5.6417	0.8824	0.7074
0.6226	188.0011	32110	6.0233	0.8836	0.7123
0.619	189.0011	32279	7.0898	0.8851	0.7149
0.6157	190.0011	32448	6.6226	0.8876	0.7232
0.6171	191.0011	32617	6.4924	0.8799	0.7070
0.6177	192.0011	32786	7.7647	0.8782	0.7098
0.6179	193.0011	32955	6.4711	0.8850	0.7151
0.6169	194.0011	33124	6.8190	0.8835	0.7149
0.6124	195.0011	33293	7.6578	0.8823	0.7119
0.6185	196.0011	33462	6.8216	0.8832	0.7175
0.6132	197.0010	33631	7.0588	0.8861	0.7242
0.6111	198.0010	33800	7.4747	0.8835	0.7144
0.6127	199.0010	33969	7.7447	0.8834	0.7113
0.61	200.0010	34138	7.6975	0.8803	0.7151
0.6193	201.0010	34307	6.4387	0.8816	0.7099
0.6167	202.0010	34476	7.5285	0.8867	0.7184
0.6157	203.0010	34645	6.6411	0.8854	0.7181
0.6082	204.0010	34814	5.7796	0.8880	0.7192
0.613	205.0010	34983	5.3015	0.8833	0.7211
0.6133	206.0010	35152	6.4275	0.8826	0.7147
0.606	207.0010	35321	6.1046	0.8792	0.7171
0.6045	208.0010	35490	5.9309	0.8844	0.7207
0.6127	209.0010	35659	6.1429	0.8845	0.7168
0.6022	210.0010	35828	6.1108	0.8857	0.7174
0.6065	211.0009	35997	6.4227	0.8778	0.7082
0.6092	212.0009	36166	6.6492	0.8842	0.7181
0.6075	213.0009	36335	5.6531	0.8819	0.7176
0.6076	214.0009	36504	6.4636	0.8832	0.7126
0.6042	215.0009	36673	6.7895	0.8773	0.7091
0.5993	216.0009	36842	7.0939	0.8837	0.7173
0.6064	217.0009	37011	6.6639	0.8863	0.7231

Framework versions

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