flan-t5-small-gec

This model is a fine-tuned version of google/flan-t5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3409
• Sacrebleu: 80.4351

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_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Sacrebleu
0.4499	0.2591	1000	0.3853	79.3662
0.4321	0.5181	2000	0.3655	79.6704
0.4061	0.7772	3000	0.3589	80.0453
0.4015	1.0363	4000	0.3518	80.1396
0.4114	1.2953	5000	0.3480	80.1401
0.4088	1.5544	6000	0.3445	80.2909
0.3998	1.8135	7000	0.3441	80.2896
0.3851	2.0725	8000	0.3429	80.3305
0.379	2.3316	9000	0.3402	80.3939
0.3756	2.5907	10000	0.3411	80.4061
0.3807	2.8497	11000	0.3409	80.4351

Framework versions

• Transformers 4.56.1
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.0