literal-support-gec

This model is a fine-tuned version of gotutiyan/gec-t5-large-clang8 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0248

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: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• 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
• lr_scheduler_warmup_steps: 500
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
0.7081	0.0373	500	0.0491
0.4451	0.0745	1000	0.0393
0.3807	0.1118	1500	0.0361
0.3942	0.1491	2000	0.0366
0.3188	0.1864	2500	0.0331
0.2964	0.2236	3000	0.0318
0.3084	0.2609	3500	0.0317
0.2826	0.2982	4000	0.0309
0.2788	0.3354	4500	0.0309
0.2838	0.3727	5000	0.0301
0.2786	0.4100	5500	0.0291
0.2707	0.4473	6000	0.0283
0.2865	0.4845	6500	0.0289
0.2580	0.5218	7000	0.0278
0.2592	0.5591	7500	0.0270
0.2332	0.5964	8000	0.0273
0.2418	0.6336	8500	0.0269
0.2305	0.6709	9000	0.0264
0.2363	0.7082	9500	0.0261
0.2385	0.7454	10000	0.0259
0.2231	0.7827	10500	0.0256
0.2227	0.8200	11000	0.0254
0.2146	0.8573	11500	0.0253
0.2261	0.8945	12000	0.0252
0.2103	0.9318	12500	0.0251
0.2178	0.9691	13000	0.0248

Framework versions

• Transformers 5.2.0
• Pytorch 2.7.0+cu126
• Datasets 4.6.1
• Tokenizers 0.22.2