flanT5_large_MT

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

• Loss: 1.9444
• Accuracy: 0.815
• Precision: 0.8339
• Recall: 0.7867
• F1 score: 0.8096

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: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1 score
1.2742	0.3910	2500	1.6400	0.69	0.7096	0.6433	0.6748
1.2294	0.7820	5000	0.9952	0.725	0.7872	0.6167	0.6916
1.0532	1.1730	7500	1.1585	0.69	0.6875	0.6967	0.6921
0.9314	1.5640	10000	0.8117	0.7417	0.7580	0.71	0.7332
0.8698	1.9550	12500	0.7872	0.7367	0.7817	0.6567	0.7138
0.7603	2.3459	15000	0.9449	0.77	0.7470	0.8167	0.7803
0.7207	2.7369	17500	0.9339	0.7917	0.7850	0.8033	0.7941
0.6447	3.1279	20000	1.0756	0.8017	0.8198	0.7733	0.7959
0.5358	3.5189	22500	1.0722	0.79	0.7788	0.81	0.7941
0.5169	3.9099	25000	1.0001	0.8067	0.8407	0.7567	0.7965
0.4011	4.3009	27500	1.0915	0.805	0.8144	0.79	0.8020
0.3177	4.6919	30000	1.2966	0.8083	0.8157	0.7967	0.8061
0.2783	5.0829	32500	1.2640	0.8117	0.7913	0.8467	0.8180
0.1352	5.4739	35000	1.3695	0.82	0.8333	0.8	0.8163
0.2168	5.8649	37500	1.3541	0.8133	0.8287	0.79	0.8089
0.1186	6.2559	40000	1.4063	0.8167	0.8276	0.8	0.8136
0.1205	6.6469	42500	1.6920	0.8033	0.7993	0.81	0.8046
0.0957	7.0378	45000	1.5681	0.815	0.8225	0.8033	0.8128
0.0406	7.4288	47500	1.9015	0.8083	0.8269	0.78	0.8027
0.0558	7.8198	50000	1.8359	0.8017	0.7967	0.81	0.8033
0.0549	8.2108	52500	1.8649	0.8083	0.8201	0.79	0.8048
0.0515	8.6018	55000	1.8556	0.8067	0.7949	0.8267	0.8105
0.0378	8.9928	57500	1.9273	0.7983	0.7806	0.83	0.8045
0.0182	9.3838	60000	1.9497	0.8133	0.8287	0.79	0.8089
0.0264	9.7748	62500	1.9444	0.815	0.8339	0.7867	0.8096

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1