segment_mask_token_v2

This model is a fine-tuned version of zera09/custom-longt5-with-ts on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0718

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	72	5.8637
21.3135	2.0	144	1.6856
1.3759	3.0	216	1.0404
1.3759	4.0	288	0.6246
0.6906	5.0	360	0.3509
0.3575	6.0	432	0.2973
0.3583	7.0	504	0.1697
0.3583	8.0	576	0.1578
0.2914	9.0	648	0.1353
0.1716	10.0	720	0.1145
0.1716	11.0	792	0.0990
0.2055	12.0	864	0.0975
0.1352	13.0	936	0.0818
0.116	14.0	1008	0.0789
0.116	15.0	1080	0.0812
0.1264	16.0	1152	0.0800
0.1599	17.0	1224	0.0762
0.1599	18.0	1296	0.0720
0.0694	19.0	1368	0.0726
0.0742	20.0	1440	0.0718

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0