HybridTimeScaleModel

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

• Loss: 3.0226

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: 1e-06
• train_batch_size: 22
• eval_batch_size: 4
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 150
• training_steps: 77000

Training results

Training Loss	Epoch	Step	Validation Loss
4.1200	0.0091	1000	4.3751
3.5707	0.0182	2000	3.8969
3.3944	0.0273	3000	3.7320
3.2534	0.0364	4000	3.6268
3.2178	0.0455	5000	3.5597
3.1700	0.0545	6000	3.4946
3.1400	0.0636	7000	3.4605
3.0755	0.0727	8000	3.4183
3.0527	0.0818	9000	3.3907
3.0213	0.0909	10000	3.3656
3.0590	0.1	11000	3.3479
2.9770	0.1091	12000	3.3242
2.9614	0.1182	13000	3.3102
2.9506	0.1273	14000	3.2945
2.9333	0.1364	15000	3.2715
2.8988	0.1455	16000	3.2674
2.9117	0.1545	17000	3.2496
2.8933	0.1636	18000	3.2416
2.8866	0.1727	19000	3.2224
2.9030	0.1818	20000	3.2188
2.9004	0.1909	21000	3.2108
2.8540	0.2	22000	3.1956
2.8207	0.2091	23000	3.1900
2.8272	0.2182	24000	3.1836
2.8192	0.2273	25000	3.1781
2.7933	0.2364	26000	3.1706
2.8126	0.2455	27000	3.1654
2.8480	0.2545	28000	3.1647
2.7938	0.2636	29000	3.1548
2.7823	0.2727	30000	3.1496
2.7479	0.2818	31000	3.1456
2.7778	0.2909	32000	3.1463
2.7698	0.0091	33000	3.1402
2.7731	0.0182	34000	3.1313
2.7693	0.0273	35000	3.1262
2.7321	0.0364	36000	3.1245
2.6874	0.0091	37000	3.1317
2.6970	0.0182	38000	3.1285
2.6958	0.0273	39000	3.1291
2.6649	0.0364	40000	3.1292
2.7598	0.0455	41000	3.1212
2.7124	0.0545	42000	3.1164
2.7572	0.0636	43000	3.1164
2.7247	0.0727	44000	3.1065
2.7101	0.0818	45000	3.1004
2.6942	0.0909	46000	3.0926
2.7299	0.1	47000	3.0938
2.7021	0.1091	48000	3.0897
2.6787	0.1182	49000	3.0872
2.6781	0.1273	50000	3.0828
2.7094	0.1364	51000	3.0797
2.7144	0.1455	52000	3.0822
2.6800	0.1545	53000	3.0733
2.6689	0.1636	54000	3.0703
2.6780	0.1727	55000	3.0706
2.6806	0.1818	56000	3.0680
2.7088	0.1909	57000	3.0626
2.6277	0.2	58000	3.0661
2.6540	0.2091	59000	3.0577
2.6756	0.2182	60000	3.0541
2.6778	0.2273	61000	3.0542
2.6535	0.2364	62000	3.0504
2.6557	0.2455	63000	3.0508
2.6445	0.2545	64000	3.0473
2.6488	0.2636	65000	3.0462
2.6457	0.2727	66000	3.0442
2.6479	0.2818	67000	3.0378
2.6303	0.2909	68000	3.0449
2.6089	0.3	69000	3.0342
2.5966	0.3091	70000	3.0347
2.6226	0.3182	71000	3.0347
2.6174	0.3273	72000	3.0310
2.6171	0.3364	73000	3.0339
2.6139	0.3455	74000	3.0277
2.5904	0.3545	75000	3.0259
2.5940	0.3636	76000	3.0242
2.5818	0.0130	77000	3.0226

Framework versions

• Transformers 5.13.1
• Pytorch 2.8.0+cu128
• Datasets 5.0.0
• Tokenizers 0.22.2