flanT5_large_Task2_semantic_pred

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: 0.9725
• Accuracy: 0.7979
• Precision: 0.8195
• Recall: 0.7566
• F1 score: 0.7868

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
1.3836	0.4074	2500	1.7839	0.5072	0.0	0.0	0.0
1.3837	0.8147	5000	0.7528	0.5072	0.0	0.0	0.0
1.2525	1.2221	7500	0.7166	0.4928	0.4928	1.0	0.6603
1.2899	1.6295	10000	1.5233	0.6310	0.5765	0.9471	0.7167
1.4214	2.0368	12500	1.2841	0.6571	0.6021	0.8968	0.7205
1.4003	2.4442	15000	1.1805	0.6115	0.5615	0.9656	0.7101
1.3396	2.8516	17500	0.9273	0.6206	0.5683	0.9577	0.7133
1.3938	3.2589	20000	1.5375	0.6858	0.7322	0.5714	0.6419
1.4272	3.6663	22500	1.2804	0.7158	0.7548	0.6270	0.6850
1.2643	4.0737	25000	1.2878	0.7262	0.7692	0.6349	0.6957
1.3015	4.4810	27500	1.1752	0.7158	0.7532	0.6296	0.6859
1.3525	4.8884	30000	1.1726	0.7249	0.6946	0.7884	0.7385
1.2839	5.2957	32500	1.0721	0.7223	0.7586	0.6402	0.6944
1.2047	5.7031	35000	1.3045	0.7223	0.7089	0.7407	0.7245
1.226	6.1105	37500	1.1262	0.7366	0.7486	0.7011	0.7240
1.1001	6.5178	40000	1.2328	0.7445	0.7407	0.7407	0.7407
1.1083	6.9252	42500	1.1119	0.7353	0.7030	0.8016	0.7491
1.0281	7.3326	45000	1.1205	0.7471	0.7434	0.7434	0.7434
1.0396	7.7399	47500	1.1243	0.7640	0.7958	0.7011	0.7454
0.9913	8.1473	50000	0.9882	0.7823	0.8187	0.7169	0.7645
0.9364	8.5547	52500	1.1489	0.7705	0.7488	0.8042	0.7755
1.0081	8.9620	55000	0.9507	0.7810	0.7807	0.7725	0.7766
0.9026	9.3694	57500	1.0181	0.7888	0.7888	0.7804	0.7846
0.8961	9.7768	60000	0.9725	0.7979	0.8195	0.7566	0.7868

Framework versions

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