t5_es_farshad_half_2_4

This model is a fine-tuned version of google-t5/t5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0456
• Accuracy: 0.9916
• F1: 0.9919

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 64
• total_train_batch_size: 4096
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.8073	5.8501	50	0.7215	0.4858	0.0155
0.659	11.7002	100	0.5497	0.8353	0.8282
0.3485	17.5503	150	0.1162	0.9684	0.9692
0.0936	23.4004	200	0.0599	0.9814	0.9821
0.0492	29.2505	250	0.0447	0.9875	0.9880
0.0316	35.1005	300	0.0426	0.9898	0.9902
0.0215	40.9506	350	0.0411	0.9890	0.9894
0.0158	46.8007	400	0.0438	0.9907	0.9911
0.0131	52.6508	450	0.0389	0.9913	0.9916
0.0108	58.5009	500	0.0352	0.9927	0.9930
0.0092	64.3510	550	0.0376	0.9922	0.9924
0.0075	70.2011	600	0.0416	0.9916	0.9919
0.0063	76.0512	650	0.0403	0.9927	0.9930
0.0052	81.9013	700	0.0426	0.9925	0.9927
0.0045	87.7514	750	0.0443	0.9919	0.9922
0.0035	93.6015	800	0.0456	0.9916	0.9919

Framework versions

• Transformers 4.40.0
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1