patchtst-tsmixup-relu

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

• Loss: 0.1485
• Mse: 229.0122
• Mae: 0.6126
• Rmse: 15.1332
• Smape: 83.2157

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: 448
• eval_batch_size: 896
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 896
• 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: 1000
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Mae	Rmse	Smape
0.1785	0.1666	1000	0.1739	433.0006	0.7291	20.8087	100.7767
0.1655	0.3333	2000	0.1654	424.8635	0.7099	20.6122	74.2051
0.1663	0.4999	3000	0.1624	381.5935	0.6823	19.5344	142.9640
0.1632	0.6666	4000	0.1599	330.2523	0.6648	18.1728	78.8419
0.162	0.8332	5000	0.1592	322.0357	0.6595	17.9454	78.0993
0.1615	0.9998	6000	0.1581	294.6714	0.6549	17.1660	79.6810
0.162	1.1665	7000	0.1573	333.3479	0.6496	18.2578	81.2097
0.1587	1.3331	8000	0.1570	263.9417	0.6429	16.2463	90.2972
0.154	1.4998	9000	0.1565	268.0528	0.6515	16.3723	83.8961
0.1562	1.6664	10000	0.1561	287.6087	0.6475	16.9590	108.2025
0.1603	1.8330	11000	0.1558	281.4109	0.6507	16.7753	77.4054
0.1557	1.9997	12000	0.1550	281.4917	0.6408	16.7777	81.8113
0.1568	2.1663	13000	0.1546	258.6138	0.6406	16.0815	89.9780
0.1556	2.3329	14000	0.1545	268.3961	0.6425	16.3828	79.0450
0.1561	2.4996	15000	0.1538	249.9753	0.6366	15.8106	88.3079
0.1546	2.6662	16000	0.1535	239.0104	0.6313	15.4600	96.2489
0.1536	2.8329	17000	0.1534	232.7196	0.6318	15.2552	73.1808
0.1531	2.9995	18000	0.1537	224.2394	0.6249	14.9746	99.7205
0.1535	3.1661	19000	0.1530	253.5844	0.6296	15.9243	80.0392
0.1532	3.3328	20000	0.1529	256.6078	0.6314	16.0190	184.8716
0.1566	3.4994	21000	0.1531	228.1704	0.6266	15.1053	90.5678
0.1547	3.6661	22000	0.1527	216.8113	0.6265	14.7245	88.1824
0.1537	3.8327	23000	0.1522	241.5133	0.6282	15.5407	73.5045
0.1531	3.9993	24000	0.1521	232.2086	0.6302	15.2384	87.4450
0.1525	4.1660	25000	0.1523	253.6224	0.6328	15.9255	88.9352
0.1525	4.3326	26000	0.1517	254.2605	0.6304	15.9455	77.5196
0.1548	4.4993	27000	0.1519	225.7644	0.6212	15.0255	82.3784
0.1527	4.6659	28000	0.1519	220.0219	0.6254	14.8331	86.0485
0.153	4.8325	29000	0.1515	258.0009	0.6347	16.0624	145.4315
0.1521	4.9992	30000	0.1516	227.8417	0.6227	15.0944	76.3474
0.151	5.1658	31000	0.1514	213.8730	0.6185	14.6244	157.4075
0.1527	5.3324	32000	0.1510	238.2835	0.6189	15.4364	571.1568
0.1529	5.4991	33000	0.1510	270.1301	0.6278	16.4356	83.5608
0.1505	5.6657	34000	0.1511	241.0177	0.6271	15.5247	76.5107
0.1521	5.8324	35000	0.1516	255.7361	0.6331	15.9918	108.9967
0.1513	5.9990	36000	0.1507	253.1635	0.6233	15.9111	85.8247
0.1502	6.1656	37000	0.1509	255.3432	0.6230	15.9795	118.9721
0.1517	6.3323	38000	0.1504	238.2068	0.6213	15.4339	79.4896
0.151	6.4989	39000	0.1508	244.4908	0.6243	15.6362	98.8420
0.1516	6.6656	40000	0.1504	229.2746	0.6231	15.1418	71.1164
0.1506	6.8322	41000	0.1501	237.0237	0.6217	15.3956	74.7138
0.1503	6.9988	42000	0.1500	240.7731	0.6206	15.5169	85.3629
0.1493	7.1655	43000	0.1501	265.2171	0.6242	16.2855	157.2147
0.1501	7.3321	44000	0.1499	247.8091	0.6219	15.7420	98.6004
0.1508	7.4988	45000	0.1497	265.8900	0.6227	16.3061	72.4383
0.1518	7.6654	46000	0.1497	249.7165	0.6216	15.8024	110.3652
0.1502	7.8320	47000	0.1496	248.1616	0.6200	15.7531	77.3612
0.1503	7.9987	48000	0.1493	237.9707	0.6190	15.4263	71.8934
0.1502	8.1653	49000	0.1494	225.7567	0.6149	15.0252	78.6202
0.1492	8.3319	50000	0.1492	258.1519	0.6185	16.0671	73.3061
0.1513	8.4986	51000	0.1491	226.3746	0.6162	15.0458	118.5835
0.1508	8.6652	52000	0.1491	236.9618	0.6171	15.3936	80.4855
0.1517	8.8319	53000	0.1490	242.2040	0.6186	15.5629	144.8560
0.1494	8.9985	54000	0.1490	237.0488	0.6174	15.3964	78.5948
0.1477	9.1651	55000	0.1488	232.7170	0.6157	15.2551	82.3074
0.1499	9.3318	56000	0.1488	236.9111	0.6168	15.3919	77.6623
0.1524	9.4984	57000	0.1487	231.8599	0.6148	15.2269	102.9215
0.1505	9.6651	58000	0.1486	230.8095	0.6139	15.1924	67.3176
0.1507	9.8317	59000	0.1485	231.5027	0.6137	15.2152	84.0686
0.1461	9.9983	60000	0.1485	229.0122	0.6126	15.1332	83.2157

Framework versions

• Transformers 4.51.3
• Pytorch 2.7.1+cu126
• Datasets 2.17.1
• Tokenizers 0.21.1