t5_es_farshad_half_2_2

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: 0.0404
• Accuracy: 0.9919
• F1: 0.9922

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
0.7201	5.8501	50	0.6804	0.6244	0.6288
0.6469	11.7002	100	0.5235	0.8538	0.8578
0.3053	17.5503	150	0.1010	0.9690	0.9695
0.0887	23.4004	200	0.0576	0.9817	0.9823
0.051	29.2505	250	0.0453	0.9869	0.9873
0.0338	35.1005	300	0.0401	0.9898	0.9902
0.0232	40.9506	350	0.0416	0.9878	0.9882
0.0165	46.8007	400	0.0401	0.9904	0.9907
0.013	52.6508	450	0.0382	0.9913	0.9916
0.0108	58.5009	500	0.0433	0.9904	0.9907
0.0089	64.3510	550	0.0363	0.9933	0.9936
0.0074	70.2011	600	0.0421	0.9913	0.9916
0.0058	76.0512	650	0.0467	0.9913	0.9916
0.005	81.9013	700	0.0446	0.9916	0.9919
0.004	87.7514	750	0.0388	0.9925	0.9927
0.0033	93.6015	800	0.0404	0.9919	0.9922

Framework versions

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