stldec_arch

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

• Loss: 0.8738

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: 128
• eval_batch_size: 128
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 512
• 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: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0	0	3.8917
2.7225	0.6483	100	2.6803
1.6189	1.2917	200	1.5324
0.9914	1.9400	300	0.9894
0.8278	2.5835	400	0.9311
0.7646	3.2269	500	0.9649
0.7273	3.8752	600	0.9046
0.706	4.5186	700	0.8855
0.6927	5.1621	800	0.8959
0.6813	5.8104	900	0.8868
0.6719	6.4538	1000	0.8772
0.6683	7.0972	1100	0.8766
0.6584	7.7455	1200	0.8761
0.6592	8.3890	1300	0.8714
0.6567	9.0324	1400	0.8729
0.6512	9.6807	1500	0.8738

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.1+cu128
• Datasets 4.4.2
• Tokenizers 0.22.1